# Banks, Firms, and Jobs

Banks, Firms, and Jobs Abstract We analyze the heterogeneous employment effects of financial shocks using a rich data set of job contracts, matched with the universe of firms and their lending banks in one Italian region. To isolate the effect of the financial shock, we construct a firm-specific time-varying measure of credit supply. The preferred estimate indicates that the average elasticity of employment to a credit supply shock is $$0.36$$. Adjustment affects both the extensive and the intensive margins and is concentrated among workers with temporary contracts. We also examine the heterogeneous effects of the credit crunch by education, age, gender and nationality. Received January 27, 2017; editorial decision December 1, 2017 by Editor Philip Strahan. Authors have furnished an Internet Appendix, which is available on the Oxford University Press Web site next to the link to the final published paper online. Introduction In the aftermath of the global financial crisis, a severe credit crunch has had long-lasting consequences on a number of advanced economies, where unemployment rates have increased markedly. The labor market effects of the crisis have not been uniform, and, in particular, young and less educated workers have been particularly hit by the crisis (Hoynes et al. 2012). Although these outcomes raise important distributional concerns, it has been argued that crises could also have a “cleansing” effect to the extent that the least productive jobs and firms are the ones relatively more affected by the financial shock (Caballero and Hammour 1994; Petrosky-Nadeau 2013). The effects of the global financial crisis on the labor market have triggered a renewed interest in the relationship between finance and employment (Pagano and Pica 2012) and, specifically, a renewed interest in the effect that credit supply shocks have on firms’ employment decisions (Chodorow-Reich 2014; Buera et al. 2015; Duygan-Bump et al. 2015). While this literature provides original insights into the effects that financial crises have on aggregate employment at the firm or state level, it is generally silent about within-firm dynamics and labor reallocation. For instance, little is known about the impact of a decline in firm financing on different types of jobs, even though a differential impact of the crisis across demographic groups would have distributional implications. Moreover, the employment adjustment within firms—between workers and jobs characterized by a different skill set—can have an effect on aggregate productivity. We contribute to this strand of literature by zooming in on the employment dynamics within the firm and by providing a series of novel findings on how firms adjust the level and composition of the labor force in response to credit shocks. In particular, we focus on worker education to test whether the contraction in employment during the global financial crisis has been associated with a skill upgrading of the workforce, at the firm level. Knowing which jobs and workers are more exposed to the real effects of large financial shocks provides useful insight into better understanding how firms reorganize themselves at times of crisis and can inform the debate on the distributional consequences and the possible cleansing effect of financial crises. We run our analysis thanks to the availability of an original and extremely rich data set, that draws on an administrative archive that collects daily information on individual job contracts and labor market flows. The data set covers the universe of firms, including micro-enterprises, in an Italian region, matched with their lending banks through the Italian Credit Register. This is an important feature of our data given that bank credit is very often the only source of external financing for micro and small enterprises. We end up with a quarterly data set of about 200,000 firms, spanning the period from 2008:Q1 to 2012:Q4, for which, thanks to the degree of granularity of the data, we can go beyond the standard job destruction/job creation dichotomy to investigate differential responses to a credit supply shocks across firms, workers, and job contracts. We find that a 10% supply-driven credit contraction reduces employment by 3.6%. This effect is the result of adjustments at the intensive and extensive margins, is concentrated among workers with temporary contracts, and mostly occurred through increased outflows rather than decreased inflows. These results are in line with the existence of a “dual” labor market in which temporary contracts absorb a large part of the employment volatility. The reduction in employment is concentrated among relatively less educated individuals and mostly happened by allowing temporary contracts to expire. By contrast, less educated workers with open-ended contracts are almost unaffected by tighter firms’ financing constraints, possibly because of higher firing costs and a rigid employment protection legislation (EPL). Even though skill upgrading strategies are heavily shaped by contract regulation, our results are not exclusively driven by high EPL: when focusing on a subsample of small firms for whom firing costs are lower, we still find that the adjustment is primarily borne by less educated workers with temporary contracts. These differential effects are mainly driven by the adjustment at the intensive margin, whereas the effects on employment due to firm exit are more homogeneous across contracts and workers. We also find evidence suggesting that women and foreign workers are hit disproportionately more by the credit shock, regardless of the kind of job contract, while the stronger effect on young workers reflects their higher propensity to be hired with temporary contracts.1 From a more general perspective, it could still be the case that workers who are more likely to lose their job in our sample could be hired by nonbanked firms. However, when we aggregate employment and loan outcomes for all firms in the region at the province-industry-quarter level, we still find a negative impact of the credit crunch (both from an economic and statistical point of view) on employment, suggesting that the contraction in employment has not been offset by firms without banking relationships. This paper contributes to the growing literature on the real effect of credit supply shocks (Amiti and Weinstein 2011, 2017; Cingano et al. 2016; Paravisini et al. 2015) and is closely related to the recent contributions that investigate the effects of financial shocks on employment outcomes at the firm level (Barbosa et al. 2017; Benmelech et al. 2015; Bentolila, Jensen, and Jimenez 2017; Berg 2016; Caggese et al. 2016; Chodorow-Reich 2014; Ersahin and Irani 2016; Giroud and Mueller 2017; Hochfellner et al. 2016; Popov and Rocholl 2016; Siemer 2016).2 Drawing on micro-level data sets, these studies consistently show that a tightening of the credit supply leads to a contraction of the workforce. The analysis by Bentolila, Jensen, and Jimenez (2017) has the unique feature of being based on loan-level data from a credit register. Relying on the differences in bank health at the beginning of the financial crisis, the paper shows that firms exposed to weak banks contracted employment by 2.8 percentage points more than firms that were borrowing from healthier lenders, and results are able to explain about a fourth of the fall in aggregate employment in Spain between 2007 and 2010. Also, their analysis uncovers that job losses have been mostly borne by temporary employees, while wages adjusted only marginally. Hochfellner et al. (2016) use employer-employee matched data for a sample of German firms to look at how individual characteristics affect labor outcomes. The identification strategy hinges on differences in firm location, distinguishing between firms that are located in one of the seven federal states in which the major bank was one of the five Landesbanks with significant exposure to the U.S. mortgage crisis and firms that are located elsewhere. In addition to confirming the aggregate negative effect of credit contraction on employment, Hochfellner et al. (2016) show that workers in firms which have been exposed to a negative credit shock experience significant earning losses and an increase in the unemployment spell. They also find that unskilled, less educated and less experienced workers are the most affected by the credit shock.3 While both these studies limit their analysis to medium-sized and large firms, Siemer (2016) uses confidential firm-level employment data from the U.S. Bureau of Labor Statistics for the universe of U.S. firms, but relies on industry-level differences in external financial dependence to identify the effects of financial constraints on employment and firm dynamics. His results show that financing constraints reduce employment growth in small firms by 5 to 10 percentage points relative to large firms, but they are silent on within-firm heterogeneity. Our analysis has the advantage of bringing together three key elements which in previous studies have been considered separately. First, the availability of loan-level data (instead of aggregate credit data) allows us to identify the bank lending channel at the firm-level. Moreover, those data make it possible to control for credit demand and productivity shocks at a granular level, with a set of firm, time, and firm cluster$$\times$$time fixed effects, which absorb firm-specific time invariant demand shifters and time-varying demand shocks that are common to a narrowly defined cluster of borrowers. The matched bank-firm data also allow us to extend the identification strategy of Greenstone et al. (2014) and construct an exogenous firm-specific time-varying measure of bank credit supply, which gives us more precise estimates than the ones obtained with more aggregate data. We start by estimating time-varying nationwide bank lending policies that are purged of local loan demand (and of any other province-sector-quarter level idiosyncratic shocks). Then we build a credit supply variable at the firm level using banks’ loan share to a given firm as weights. We discuss different arguments to motivate the exogeneity of our instrument and we show that it is strongly correlated with loan growth at the firm level.4 Second, thanks to contract-firm-bank-matched data, we can investigate heterogeneous responses to a financial shock across workers, job contracts, and firms. In particular, we can exploit differences across contract types and look at the intersection between demographic characteristics (education, age, gender and nationality) and job contracts, to assess which dimensions matter more for firm’s employment decisions.5 Third and finally, our analysis covers the universe of firms. While there is a wide consensus on the fact that smaller firms rely more on bank financing, the existing evidence rarely focuses on a representative sample of small firms. Our data, on the contrary, include the universe of individual and micro enterprises, and this allows us to have a more precise (and larger) estimate of the employment effect of financial shocks. 1. Data 1.1 Veneto as a representative case study Our analysis relies on unparalleled loan-level information about the entire population of workers, firms and financial intermediaries operating in Veneto, a large Italian region with a population of 4.9 million individuals and a workforce of 2.2 million workers. According to the National Institute of Statistics data, the region accounts for roughly 9% of the Italian value added and of total employment. A key feature for our analysis is that Veneto can be considered as a self-contained labor market. About 97% of the workers who reside in the region have their workplace in a municipality within the region, and migration to other regions is a negligible phenomenon at the aggregate level (0.4% of the population per year); moreover, both figures are substantially stable in the temporal window considered in the analysis. As a result, it is unlikely that our results will be biased by dismissed workers finding jobs out of region. Veneto shares with Italy a large prevalence of small firms (Figure 1, panel a): 94% of firms in the region have fewer than 10 employees (57% have at most one employee). The productive structure is also fairly similar to the national one (Figure 1, panel b), and the service and industrial sectors accounts for 56% and 43% of total employment, respectively, with the share of the industrial sector being slightly larger than in the rest of Italy. Figure 1 View largeDownload slide External validity: Firm distribution across size and sectors in Veneto and Italy The figure further elaborates ISTAT data (census 2011). Figure 1 View largeDownload slide External validity: Firm distribution across size and sectors in Veneto and Italy The figure further elaborates ISTAT data (census 2011). In terms of the banking system, in 2012 in Veneto there were about 120 banks, with small local banks accounting for nearly 20% of business loans. The degree of financial development, as measured by the number of branches per inhabitants, is higher with respect to the national average (Figure 2, panel a). Aggregate lending to nonfinancial corporations followed a similar dynamic in Veneto and Italy (Figure 2, panel b). Figure 2 View largeDownload slide External validity: Bank penetration and lending in Veneto and Italy The figure further elaborates data from Bank of Italy. Figure 2 View largeDownload slide External validity: Bank penetration and lending in Veneto and Italy The figure further elaborates data from Bank of Italy. Veneto is hence very well representative of the Italian situation, which in turn represents an extremely interesting case studies for at least two reasons: first, Italian firms mostly rely on bank credit for their business activities, and more than other firms in the Euro area (Figure 3, panel a); second, small firms (fewer than 10 employees) are the most indebted, and the Italian productive structure is strongly biased toward small production units (Figure 3, panel b). Figure 3 View largeDownload slide Bank financing in Italy across firm size The figure further elaborates data from the Survey on the Access to Finance of Enterprises (SAFE, European Central Bank), Bank of Italy, PLANET, and ASIA. Debt per employee is measured in thousands of euro. Figure 3 View largeDownload slide Bank financing in Italy across firm size The figure further elaborates data from the Survey on the Access to Finance of Enterprises (SAFE, European Central Bank), Bank of Italy, PLANET, and ASIA. Debt per employee is measured in thousands of euro. 1.2 The contract-firm-bank-matched data Our data set brings together an extremely rich set of information coming from different administrative sources. In the following we provide an overview of the construction and structure of data set, while more detailed information are discussed in the annex A-I. Daily labor market flows from the regional public employment service are indeed matched to stock information from the national social security administration and to the Italian credit register maintained by the Bank of Italy using firm-level unique identifiers, namely their VAT numbers. These feature of the data guarantees at the same time wide population coverage, high information reliability and a nearly total frequency of success in the matching procedure. The bulk of labor market information comes from PLANET, an administrative data set of daily labor market flows maintained by the regional employment agency Veneto Lavoro. PLANET builds on the obligation for firms operating in Italy to notice the national and local employment agencies about all labor market transitions for which they are held responsible, including hires, firings and transformations of individual employment arrangements (e.g., from full-time to part-time, from temporary to permanent, and the like). Firm-level observables include geographical location and sector (5-digit NACE code), while worker information covers gender, age, nationality, occupation (5-digit ISCO code), type of contract (44 different employment arrangement), educational attainment (13 categories), time schedule (full-time or vertical, horizontal or mixed part-time), and reasons for separation from the firm. To overcome limitations in terms of labor market stocks, PLANET is complemented with information from ASIA, the archive of active firms maintained by the National Statistical Institute (ISTAT) with register data from the Social Security Administration. ASIA provides yearly data about firms whose economic activity spans for at least six months within a calendar year. To our purposes, ASIA adds information on firm size and on characteristics of those firms who are not interested by any job flows or transitions in our sample period. More specifically, we consider the stock in the first year in which we observe the firm, and we reconstruct the stock forward using information on workers inflows and outflows. The purpose of this exercise is to guarantee consistency between flows and stocks and, more importantly, to have quarterly stock data.6 To obtain a firm-specific measure of credit availability, we use information from the Credit Register (CR) database, managed by the Bank of Italy, on the credit extended to each firm in each quarter. For each borrower, banks have to report to the Register, on a monthly basis, the amount of each loan—granted and used—for all loans exceeding a minimum threshold (75,000 euro until December 2008, 30,000 euro afterward), plus all nonperforming loans. Given the low threshold, these data can be taken as a census.7 Data also contain a breakdown by type of the loan (e.g., credit lines, credit receivables and fixed-term loans). From CR we essentially draw two kind of information. First, borrower’s outstanding loans (from all banks operating in Italy) at the end of each quarter: we consider the total amount instead of the different types of loans because banks and borrowers may endogenously change the composition of loans in reaction to shocks to the credit market. Second, the bank market share for each borrower at the beginning of the period, that we use to construct the instrumental variable (see Section 2.2).8 1.3 Sample selection and the final data set All data sources are merged together using VAT numbers as univocal firm identifiers. Genuine nonmatches between PLANET and ASIA are possible, and are due to two reasons: very short-lived firms (less than a semester in a calendar year) are not recorded in ASIA, while firms with a very stable employed workforce (meaning no changes in both the intensive and the extensive margins, including the type of contract) do not appear in PLANET. None of the two entails any limitation to our purposes, as (1) the stock of employed workforce for very short-lived firms can be easily induced from workers’ flows, and (2) the worker flows in stable firms are by definition null. Moreover, all firms with loan information are also present in ASIA, so extremely short-lived firms fall beyond the scope of the analysis. Thus, we include all firms that are not in PLANET but are in the firm register, and we assume that inflows and outflows for those firms are zero. This grants that truly unsuccessful matches are infrequent and largely due to misreporting of VAT numbers by either the firms or the statistical offices maintaining the single sources, an occurrence that we can safely assume to be random and—due to the extremely large sample size—almost irrelevant from a statistical standpoint. The selection of the sample is driven by two main reasons. First, although the available time series cover a longer period, we narrow our focus on the years from 2008 to 2012 (the last available year in most sources at the time of our analysis). The reason is that until 2007 the obligation for firms to notice hires and firings (from which PLANET originates) concerned dependent workers only and occurred largely through paper documents. The first limitation resulted in an incomplete coverage of labor market flows, insofar as independent contractors and disguised self-employees—widely spread in the Italian labor market and at high risk to represent a buffer stock of employment during downturns—were not observed in the data. The second limitation entailed a nonnegligible delay of data completion. Both were overcome during 2007, when digital notice became compulsory for all workers, including independent ones. Second, we focus on the private nonfinancial nonprimary sectors. The reasons are self-evident. Employment in the public sector depends on different rationales that include macroeconomic stabilization, budget control and the supply of public services, and its funding relies to a great deal on out-of-market sources (taxes). The agriculture sector in turn is highly subsidized all over the EU and a credit crunch from the private sector may be overcome by financial resources that we cannot observe at the micro level. Finally, credit flows within the financial sector often respond to different factors than flows from banks to nonfinancial corporations.9 After a process of data cleansing, the final sample includes nearly 440,000 firms of which about 200,000 have bank relationships. 1.4 Descriptive statistics The firms included in the sample are predominantly micro and small enterprises, reflecting the structure of the Italian industry. This distribution is consistent with Census data both in terms of firms and employees (Figure 4). Over the sample period 2008–2012, the number of employees declines by nearly 90,000 units, and the number of firms records a significant drop too. These trends mimic the aggregate data from the National Institute of Statistics (Figure 5). Figure 4 View largeDownload slide A comparison with the Census data: Sample representativeness The figure further elaborates data from ISTAT (2011 census), PLANET, and ASIA. Figure 4 View largeDownload slide A comparison with the Census data: Sample representativeness The figure further elaborates data from ISTAT (2011 census), PLANET, and ASIA. Figure 5 View largeDownload slide Dynamics of firms and employment: Sample representativeness The figure further elaborates data from PLANET, ASIA and ISTAT (“Labour Force Survey”). Data on the total economy (panel b) come from ISTAT and are computed filtering the data along a sectoral composition as close as possible to that of the firms included in the sample: private nonfinancial nonprimary sectors. Figure 5 View largeDownload slide Dynamics of firms and employment: Sample representativeness The figure further elaborates data from PLANET, ASIA and ISTAT (“Labour Force Survey”). Data on the total economy (panel b) come from ISTAT and are computed filtering the data along a sectoral composition as close as possible to that of the firms included in the sample: private nonfinancial nonprimary sectors. Temporary contracts, which account for more than 10% of all contracts (Table 1), could act as a buffer for firms to adjust to a credit shock in the very short term. The average duration of temporary contracts in our sample is 9.4 months, and about two thirds of the temporary contracts end within a quarter. Table 1 Summary statistics Variable Mean SD Share in total employment (%) $$\Delta Employment$$ - Total $$-$$0.0211 0.2610 100.0 $$\quad$$Low education $$-$$0.0098 0.1480 38.7 $$\quad$$Medium education $$-$$0.0088 0.1420 48.0 $$\quad$$High education $$-$$0.0022 0.0569 13.3 $$\quad$$Under 30 $$-$$0.0025 0.1080 17.9 $$\quad$$Over 30 $$-$$0.0186 0.2090 82.1 $$\quad$$Male $$-$$0.0126 0.1690 59.9 $$\quad$$Female $$-$$0.0085 0.1470 40.1 $$\quad$$Italian $$-$$0.0179 0.2260 91.4 $$\quad$$Foreign $$-$$0.0033 0.0919 8.6 $$\Delta Employment$$ - Open-ended $$-$$0.0173 0.1890 88.7 $$\quad$$Low education $$-$$0.0080 0.1000 32.8 $$\quad$$Medium education $$-$$0.0075 0.1020 45.0 $$\quad$$High education $$-$$0.0016 0.0335 11.0 $$\quad$$Under 30 $$-$$0.0015 0.0533 15.0 $$\quad$$Over 30 $$-$$0.0158 0.1620 73.8 $$\quad$$Male $$-$$0.0102 0.1220 51.0 $$\quad$$Female $$-$$0.0071 0.0984 37.7 $$\quad$$Italian $$-$$0.0148 0.1660 80.1 $$\quad$$Foreign $$-$$0.0025 0.0596 8.6 $$\Delta Employment$$ - Temporary $$-$$0.0039 0.1660 11.3 $$\quad$$Low education $$-$$0.0016 0.1060 3.9 $$\quad$$Medium education $$-$$0.0017 0.0993 5.4 $$\quad$$High education $$-$$0.0006 0.0464 2.0 $$\quad$$Under 30 $$-$$0.0014 0.0972 5.4 $$\quad$$Over 30 $$-$$0.0025 0.1210 5.9 $$\quad$$Male $$-$$0.0019 0.1080 4.9 $$\quad$$Female $$-$$0.0019 0.1120 6.3 $$\quad$$Italian $$-$$0.0029 0.1410 9.8 $$\quad$$Foreign $$-$$0.0009 0.0717 1.5 $$\Delta Loan$$ $$-$$0.0163 0.3151 – $$CSI$$ $$-$$0.0085 0.0404 – $$EXIT$$ 0.0066 0.0811 – Variable Mean SD Share in total employment (%) $$\Delta Employment$$ - Total $$-$$0.0211 0.2610 100.0 $$\quad$$Low education $$-$$0.0098 0.1480 38.7 $$\quad$$Medium education $$-$$0.0088 0.1420 48.0 $$\quad$$High education $$-$$0.0022 0.0569 13.3 $$\quad$$Under 30 $$-$$0.0025 0.1080 17.9 $$\quad$$Over 30 $$-$$0.0186 0.2090 82.1 $$\quad$$Male $$-$$0.0126 0.1690 59.9 $$\quad$$Female $$-$$0.0085 0.1470 40.1 $$\quad$$Italian $$-$$0.0179 0.2260 91.4 $$\quad$$Foreign $$-$$0.0033 0.0919 8.6 $$\Delta Employment$$ - Open-ended $$-$$0.0173 0.1890 88.7 $$\quad$$Low education $$-$$0.0080 0.1000 32.8 $$\quad$$Medium education $$-$$0.0075 0.1020 45.0 $$\quad$$High education $$-$$0.0016 0.0335 11.0 $$\quad$$Under 30 $$-$$0.0015 0.0533 15.0 $$\quad$$Over 30 $$-$$0.0158 0.1620 73.8 $$\quad$$Male $$-$$0.0102 0.1220 51.0 $$\quad$$Female $$-$$0.0071 0.0984 37.7 $$\quad$$Italian $$-$$0.0148 0.1660 80.1 $$\quad$$Foreign $$-$$0.0025 0.0596 8.6 $$\Delta Employment$$ - Temporary $$-$$0.0039 0.1660 11.3 $$\quad$$Low education $$-$$0.0016 0.1060 3.9 $$\quad$$Medium education $$-$$0.0017 0.0993 5.4 $$\quad$$High education $$-$$0.0006 0.0464 2.0 $$\quad$$Under 30 $$-$$0.0014 0.0972 5.4 $$\quad$$Over 30 $$-$$0.0025 0.1210 5.9 $$\quad$$Male $$-$$0.0019 0.1080 4.9 $$\quad$$Female $$-$$0.0019 0.1120 6.3 $$\quad$$Italian $$-$$0.0029 0.1410 9.8 $$\quad$$Foreign $$-$$0.0009 0.0717 1.5 $$\Delta Loan$$ $$-$$0.0163 0.3151 – $$CSI$$ $$-$$0.0085 0.0404 – $$EXIT$$ 0.0066 0.0811 – The table reports the summary statistics for: (1) $$\Delta Employment$$ for different demographic characteristics, for all contracts and separately for open-ended and temporary contracts; (2) the average change in firm borrowing over two quarters ($$\Delta Loan$$); (3) the credit supply index ($$CSI$$); and (4) a binary variable identifying firms that closed their activity in a given quarter $$t$$, but were active in $$t-1$$ ($$EXIT$$). The sample is the one used in the empirical analysis, made by the universe of firms, conditional on having bank debt. The change in employment for temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. Education level are defined as low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education, based on the ISCED classification. The last column report the share of employment at the beginning of the period (end 2007) for different characteristics of contract and workers: these data are taken from the “Labour Force Survey” of the National Institute of Statistics. Table 1 Summary statistics Variable Mean SD Share in total employment (%) $$\Delta Employment$$ - Total $$-$$0.0211 0.2610 100.0 $$\quad$$Low education $$-$$0.0098 0.1480 38.7 $$\quad$$Medium education $$-$$0.0088 0.1420 48.0 $$\quad$$High education $$-$$0.0022 0.0569 13.3 $$\quad$$Under 30 $$-$$0.0025 0.1080 17.9 $$\quad$$Over 30 $$-$$0.0186 0.2090 82.1 $$\quad$$Male $$-$$0.0126 0.1690 59.9 $$\quad$$Female $$-$$0.0085 0.1470 40.1 $$\quad$$Italian $$-$$0.0179 0.2260 91.4 $$\quad$$Foreign $$-$$0.0033 0.0919 8.6 $$\Delta Employment$$ - Open-ended $$-$$0.0173 0.1890 88.7 $$\quad$$Low education $$-$$0.0080 0.1000 32.8 $$\quad$$Medium education $$-$$0.0075 0.1020 45.0 $$\quad$$High education $$-$$0.0016 0.0335 11.0 $$\quad$$Under 30 $$-$$0.0015 0.0533 15.0 $$\quad$$Over 30 $$-$$0.0158 0.1620 73.8 $$\quad$$Male $$-$$0.0102 0.1220 51.0 $$\quad$$Female $$-$$0.0071 0.0984 37.7 $$\quad$$Italian $$-$$0.0148 0.1660 80.1 $$\quad$$Foreign $$-$$0.0025 0.0596 8.6 $$\Delta Employment$$ - Temporary $$-$$0.0039 0.1660 11.3 $$\quad$$Low education $$-$$0.0016 0.1060 3.9 $$\quad$$Medium education $$-$$0.0017 0.0993 5.4 $$\quad$$High education $$-$$0.0006 0.0464 2.0 $$\quad$$Under 30 $$-$$0.0014 0.0972 5.4 $$\quad$$Over 30 $$-$$0.0025 0.1210 5.9 $$\quad$$Male $$-$$0.0019 0.1080 4.9 $$\quad$$Female $$-$$0.0019 0.1120 6.3 $$\quad$$Italian $$-$$0.0029 0.1410 9.8 $$\quad$$Foreign $$-$$0.0009 0.0717 1.5 $$\Delta Loan$$ $$-$$0.0163 0.3151 – $$CSI$$ $$-$$0.0085 0.0404 – $$EXIT$$ 0.0066 0.0811 – Variable Mean SD Share in total employment (%) $$\Delta Employment$$ - Total $$-$$0.0211 0.2610 100.0 $$\quad$$Low education $$-$$0.0098 0.1480 38.7 $$\quad$$Medium education $$-$$0.0088 0.1420 48.0 $$\quad$$High education $$-$$0.0022 0.0569 13.3 $$\quad$$Under 30 $$-$$0.0025 0.1080 17.9 $$\quad$$Over 30 $$-$$0.0186 0.2090 82.1 $$\quad$$Male $$-$$0.0126 0.1690 59.9 $$\quad$$Female $$-$$0.0085 0.1470 40.1 $$\quad$$Italian $$-$$0.0179 0.2260 91.4 $$\quad$$Foreign $$-$$0.0033 0.0919 8.6 $$\Delta Employment$$ - Open-ended $$-$$0.0173 0.1890 88.7 $$\quad$$Low education $$-$$0.0080 0.1000 32.8 $$\quad$$Medium education $$-$$0.0075 0.1020 45.0 $$\quad$$High education $$-$$0.0016 0.0335 11.0 $$\quad$$Under 30 $$-$$0.0015 0.0533 15.0 $$\quad$$Over 30 $$-$$0.0158 0.1620 73.8 $$\quad$$Male $$-$$0.0102 0.1220 51.0 $$\quad$$Female $$-$$0.0071 0.0984 37.7 $$\quad$$Italian $$-$$0.0148 0.1660 80.1 $$\quad$$Foreign $$-$$0.0025 0.0596 8.6 $$\Delta Employment$$ - Temporary $$-$$0.0039 0.1660 11.3 $$\quad$$Low education $$-$$0.0016 0.1060 3.9 $$\quad$$Medium education $$-$$0.0017 0.0993 5.4 $$\quad$$High education $$-$$0.0006 0.0464 2.0 $$\quad$$Under 30 $$-$$0.0014 0.0972 5.4 $$\quad$$Over 30 $$-$$0.0025 0.1210 5.9 $$\quad$$Male $$-$$0.0019 0.1080 4.9 $$\quad$$Female $$-$$0.0019 0.1120 6.3 $$\quad$$Italian $$-$$0.0029 0.1410 9.8 $$\quad$$Foreign $$-$$0.0009 0.0717 1.5 $$\Delta Loan$$ $$-$$0.0163 0.3151 – $$CSI$$ $$-$$0.0085 0.0404 – $$EXIT$$ 0.0066 0.0811 – The table reports the summary statistics for: (1) $$\Delta Employment$$ for different demographic characteristics, for all contracts and separately for open-ended and temporary contracts; (2) the average change in firm borrowing over two quarters ($$\Delta Loan$$); (3) the credit supply index ($$CSI$$); and (4) a binary variable identifying firms that closed their activity in a given quarter $$t$$, but were active in $$t-1$$ ($$EXIT$$). The sample is the one used in the empirical analysis, made by the universe of firms, conditional on having bank debt. The change in employment for temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. Education level are defined as low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education, based on the ISCED classification. The last column report the share of employment at the beginning of the period (end 2007) for different characteristics of contract and workers: these data are taken from the “Labour Force Survey” of the National Institute of Statistics. Looking at the subsample of the indebted firms (i.e., those used in the empirical analysis), the average firm has 6.3 employees (the median is 2 employees); two-thirds of the firms are in the service sector. In terms of the geographical distribution, firms are roughly equally distributed across the seven provinces of Veneto, with Padua (20%) and Verona (19%) being the two most populated provinces, and Venice (the regional capital) accounting for 16% of firms. Finally, our sample mostly includes firms that borrow from one bank, while about one-third of firms were borrowing from more than one bank at the beginning of the period. The job loss for the average firm is equal to 2.1%, while credit declined by 1.6% (see Table 1) consistent with the evidence of a significant credit crunch in Italy following the Lehman’s collapse (Presbitero et al. 2014; Cingano et al. 2016).10 However, the reduction in bank credit and employment was heterogeneous, as one fourth of firms experienced a negative change in employment and credit contracted for more than half of the firms in the sample. 2. Identification Strategy 2.1 The empirical model We test for the effect of credit supply on firm employment decisions estimating the following model: \begin{align}\label{eq:eq1} \Delta Employment_{it} =\beta \Delta Loan_{it} + \delta_{i} + (\gamma_{s(i)} \times \tau_t) + (\eta_{c(i)} \times \tau_t) + (\theta_{p(i)} \times \tau_t) + \epsilon_{it} , \end{align} (1) where the changes in total employment ($$\Delta Employment_{it}$$) and in loans used by the banking system $$\Delta Loan_{it}$$ for firm $$i$$ over the quarter $$t$$, are calculated as \begin{align}\label{eq:delta} \Delta X_{it} =\frac{X_{it_1} - X_{it_0}}{0.5 \times X_{it_1} + 0.5 \times X_{it_0}}, \end{align} (2) where $$X_{t_0}$$ and $$X_{t_1}$$ are, respectively, the values of employment and bank lending at the beginning and the end of the quarter $$t$$. Variations calculated in this way are widely used because they have the advantage of being symmetric and bounded between $$-2$$ (exiters) and $$+2$$ (entrants) and they are equal to zero for firms that do not register any variation in employment or lending within the quarter (Moscarini and Postel-Vinay 2012; Haltiwanger et al. 2013; Siemer 2016).11 Since labor decisions are sticky and the real effects of a financial shock could be visible with some lag (Greenstone et al. 2014; Popov and Rocholl 2016), in the baseline specification we consider the average change in used loans over two quarters (formally, we calculate $$\Delta Loan_{it}$$ and $$\Delta Loan_{it-1}$$ and we take the average change).12 Summary statistics for these variables—for different job contracts and workers—are reported in Table 1. The estimate of $$\beta$$ gives the magnitude of the bank lending channel on employment dynamics. To assess the effect of bank lending on firm employment we face two main challenges. First, the observed amount of bank credit is the equilibrium of demand for and supply of credit. To deal with possible demand and productivity shocks we first add firm and time (quarter) effects, which allow for firm-specific time invariant demand shifters and for common global shocks occurring at a quarterly frequency. Then we saturate the model with more sophisticated (2-digit) industry$$\times$$quarter ($$\gamma_{s(i)} \times \tau_t$$) and province$$\times$$quarter ($$\theta_{p(i)} \times \tau_t$$) fixed effects, and with a set of dummies that vary across quarters and firm class size (micro, small and medium-large firms, $$\eta_{c(i)} \times \tau_t$$). The degree of granularity of these borrower fixed effects is such that our identification hinges on the assumptions that (1) firm unobserved heterogeneity that drives labor demand (i.e., managerial risk appetite) is time invariant and (2) all firms operating in the same 2-digit industry, in the same province, and in the same class size face the same demand or productivity shock in each quarter. Given that we consider the universe of firms in a relatively homogeneous region, we believe that such granular fixed effects should be sufficient to isolate time-varying unobserved demand shocks. That said, we run additional robustness test allowing for more demanding firm cluster$$\times$$time fixed effects to absorb time-varying borrower demand shocks, using industry-province-size-quarter fixed effects (see Section 5). Second, bank lending is endogenous to firms’ economic conditions and employment choices, so that standard ordinary least squares (OLS) estimates are likely to be biased.13 To isolate a credit supply shock from a lower demand for credit we build on an instrumental variable (IV) approach similar to the one proposed by Greenstone et al. (2014). We construct a time-varying firm-specific index of credit supply ($$CSI_{it}$$)—discussed in detail in the following section—and we use it as an instrument for $$\Delta Loan_{it}$$. In this way, we can measure the firm-level “aggregate” bank lending channel (Jiménez et al. 2014), which takes into account general equilibrium effects (i.e., the possibility that firms substitute for credit across banks). 2.2 Credit supply index To isolate the exogenous component of credit supply, we adopt a data-driven approach, in the spirit of Greenstone et al. (2014). Specifically, we estimate the following equation that decomposes the contribution of demand and supply factors to bank lending growth at the national level: \begin{align}\label{eq:bank} \Delta L_{bpst} = \alpha + \delta_{bt} + \gamma_{pst} + \epsilon_{bpst}, \end{align} (3) where the outcome variable $$\Delta L_{bpst}$$ is the percentage change in outstanding business loans by bank $$b$$, in province $$p$$, in sector $$s$$ at time $$t$$; specifically we observe outstanding loans for about 650 banks, 100 provinces (after excluding those located in Veneto), and the main sectors of activity (agriculture, manufacturing, construction, and private nonfinancial services);14$$\gamma_{pst}$$ is a set of province-sector-quarter fixed effects that capture the variation in the change of lending due to province-sector cycles, which can be interpreted as broadly measuring local demand; and the bank-time fixed effects $$\delta_{bt}$$ represent our parameters of interest and capture (nationwide) bank lending policies. The identification of both $$\gamma_{pst}$$ and $$\delta_{bt}$$ is guaranteed by the presence of multiple banks in each province-sector market (i.e., multiple banks exposed to the same demand) and the presence of each bank in multiple province-sector markets (i.e., multiple markets exposed to the same bank supply conditions). We then construct a time-varying firm-specific index of credit supply, aggregating the bank-specific supply shocks estimated above with the beginning-of-the-period banks’ shares at the firm level as weights. Specifically, the credit supply for the firm $$i$$ at time $$t$$ is \begin{align}\label{eq:CSI} CSI_{it} = \sum_{b} w_{bit_0} \times \hat{\delta}_{bt}, \end{align} (4) where $$\hat{\delta}_{bt}$$ are the bank-time fixed effects estimated in Equation (3) and $$w_{bit_0}$$ is the bank $$b$$ market share for firm $$i$$ at the beginning of the sample period (end-2007). By construction, $$CSI_{it}$$ captures the time-varying credit supply at the firm level and its sources of variability are the substantial heterogeneity in changes in business lending across banks and the variation in bank market shares across firms. To further convince the reader that our measure of credit supply is actually correlated with the evolution of credit conditions in Italy and with bank characteristics, we provide a set of stylized facts. First, we show that, at the nationwide level, the evolution of bank lending policies mimics quite well the growth rate of business loans; the correlation is stronger in the first part of the crisis and weaker in more recent years (Figure 6, panel a); and the latter pattern might be due to the prevalence of demand factors in the second part of the crisis as main drivers of loan growth rate. More interestingly, from a microeconomic point of view, banks applying different conditions in terms of access to credit are characterized by significant differences in loans dynamics. Specifically, for each period we divide banks into two groups, depending on whether their estimated credit supply orientation ($$\hat{\delta}_{bt}$$) was below or above the median, and we examine credit patterns for both groups: as expected, tight banks recorded more negative patterns than ease ones (Figure 6, panel b). Next, we can see that there is significant variability in credit supply across banks, with the large contraction in the supply of credit around 2009 being driven by banks with the lowest values of $$\hat{\delta}_{bt}$$ (Figure 6, panel c).15 Finally, the time pattern of our credit supply indicator is also consistent with other aggregate indicators measuring the credit supply orientation. Specifically, in panel d) of Figure 6 we plot the (inverse of) $$CSI$$ together with (1) the diffusion index from the ECB Bank Lending Survey on Italian banks,16 (2) the share of rationed firms as reported by a survey on firms maintained by the Bank of Italy, and (3) a corporate credit rationing indicator developed by Burlon et al. (2016) using bank-firm-matched data. The chart shows that the credit supply index closely follows the evolution of bank lending standards and the ones of firm financing constraints; the correlation of the $$CSI$$ with the three measures of credit constraints varies between $$0.6$$ and $$0.7$$. Figure 6 View largeDownload slide View largeDownload slide Bank lending policies and credit supply index: Descriptive statistics The time-varying nationwide bank lending policies ($$\hat{\delta}_{bt}$$) at the bank level and the credit supply index ($$CSI_{it}$$) at the firm level are obtained following the approach by Greenstone et al. (2014), as discussed in Section 2.2 (see specifically Equations (3) and (4), respectively). The credit supply index is constructed aggregating the bank-quarterly fixed effects ($$\hat{\delta}_{bt}$$) with initial banks’ market share. All charts refer to Italy. Panel (a) reports the average bank lending policy obtained averaging the bank-level $$\hat{\delta}_{bt}$$ weighted by bank market share in terms of loans. In panel (b), tight (ease) banks are those that, in each quarter, have a bank lending policy ($$\hat{\delta}_{bt}$$) below (above) the median. In panel (c) we divide banks depending their lending policies ($$\hat{\delta}_{bt}$$) and we report the evolution of $$\hat{\delta}_{bt}$$ for banks at the 25th, 50th and 75th percentiles. Panel (d) plots four indicators, all standardized to make the comparison easier: (a) the inverse of the $$CSI$$; (b) the Diffusion index, calculated from answers to question 1 (“Over the past 3 months, how have your bank’s credit standards as applied to the approval of loans or credit lines to enterprises changed?”) of the ECB Bank Lending Survey on Italian Banks (the five possible answers to questions 1 and 6 are: (1) tighten considerably, (2) tighten somewhat, (3) remain basically unchanged, (4) ease somewhat, and (5) ease considerably. The diffusion index varies between -1 and 1; it is computed as the weighted mean of answers (1)–(5), where the values attributed to each answer are 1, 0.5, 0, -0.5, and -1, and the weights are the observed frequencies. See www.ecb.int/stats/money/surveys/lend/html/index.en.html); (c) the share of rationed firms as reported by a survey on firms maintained by the Bank of Italy (INVIND): firms are considered as credit constrained if they asked banks or other financial intermediaries for more credit, and the request has been denied (even in part); and (d) a measure of corporate credit rationing: Burlon et al. (2016) identifies whether or not any bank-firm transaction is credit rationed or not through the estimation of supply and demand curves and under the assumption that the observed quantity of credit is the minimum between the demand and supplied quantities. Source: Elaboration on data drawn from the Bank of Italy SR, CR, BLS, INVIND; European Central Bank; and Burlon et al. (2016) Figure 6 View largeDownload slide View largeDownload slide Bank lending policies and credit supply index: Descriptive statistics The time-varying nationwide bank lending policies ($$\hat{\delta}_{bt}$$) at the bank level and the credit supply index ($$CSI_{it}$$) at the firm level are obtained following the approach by Greenstone et al. (2014), as discussed in Section 2.2 (see specifically Equations (3) and (4), respectively). The credit supply index is constructed aggregating the bank-quarterly fixed effects ($$\hat{\delta}_{bt}$$) with initial banks’ market share. All charts refer to Italy. Panel (a) reports the average bank lending policy obtained averaging the bank-level $$\hat{\delta}_{bt}$$ weighted by bank market share in terms of loans. In panel (b), tight (ease) banks are those that, in each quarter, have a bank lending policy ($$\hat{\delta}_{bt}$$) below (above) the median. In panel (c) we divide banks depending their lending policies ($$\hat{\delta}_{bt}$$) and we report the evolution of $$\hat{\delta}_{bt}$$ for banks at the 25th, 50th and 75th percentiles. Panel (d) plots four indicators, all standardized to make the comparison easier: (a) the inverse of the $$CSI$$; (b) the Diffusion index, calculated from answers to question 1 (“Over the past 3 months, how have your bank’s credit standards as applied to the approval of loans or credit lines to enterprises changed?”) of the ECB Bank Lending Survey on Italian Banks (the five possible answers to questions 1 and 6 are: (1) tighten considerably, (2) tighten somewhat, (3) remain basically unchanged, (4) ease somewhat, and (5) ease considerably. The diffusion index varies between -1 and 1; it is computed as the weighted mean of answers (1)–(5), where the values attributed to each answer are 1, 0.5, 0, -0.5, and -1, and the weights are the observed frequencies. See www.ecb.int/stats/money/surveys/lend/html/index.en.html); (c) the share of rationed firms as reported by a survey on firms maintained by the Bank of Italy (INVIND): firms are considered as credit constrained if they asked banks or other financial intermediaries for more credit, and the request has been denied (even in part); and (d) a measure of corporate credit rationing: Burlon et al. (2016) identifies whether or not any bank-firm transaction is credit rationed or not through the estimation of supply and demand curves and under the assumption that the observed quantity of credit is the minimum between the demand and supplied quantities. Source: Elaboration on data drawn from the Bank of Italy SR, CR, BLS, INVIND; European Central Bank; and Burlon et al. (2016) Second, our measure of credit supply shows the expected correlation with bank characteristics. We run a set of bank-level regressions on the cross-section of banks, taking the average of individual nationwide bank lending policies $$\hat{\delta}_{bt}$$ over the period 2008-2012 as the dependent variables and a set of bank characteristics measured at end-2007 as explanatory variables. The worsening in credit supply conditions was higher for larger banks and those with larger funding gap (measured with the deposit-to-loan ratio) and with lower capital, consistent with the fact that those banks were likely more exposed to the liquidity drought in interbank markets and, more generally, to the financial turmoil (see Online Appendix Table A1). The exogeneity of $$CSI_{it}$$ relies on the two terms $$w_{bit_0}$$ and $$\hat{\delta}_{bt}$$. As for the first term, our assumption is that the bank market shares at the firm level, once we have controlled for firm-fixed effects, are not correlated with the employment trend at the firm level. Though this is a reasonable assumption, one may still have some concerns. For instance, one could think that bank business model may play a role. In that case, large banks could specialize in lending to large firms that are more exposed to the economic cycle (thus experiencing a decrease in employment) and if those same banks also restricted credit supply more than other players, then a correlation between our credit supply indicator and firm employment growth would be spurious. To address this issue, we include in the specification industry$$\times$$quarter and class size$$\times$$quarter fixed effects. As our parameter of interest ($$\beta$$ in Equation (1)) is fairly stable (see Section 3.1), we argue that the problem discussed above is not likely to be an issue in our case. Moreover, as shown in Table 2 on balancing properties, the exposure to credit shocks at the firm level in our sample period (obtained averaging $$CSI_{it}$$ over the period 2008–2012) is not significantly correlated (both from a statistical and economic point of view) to firm size at the beginning-of-the-period. Table 2 Orthogonality conditions Quintile of exposure to credit supply shock Correlation with 1 2 3 4 5 credit supply ($$CSI$$) Credit supply index ($$CSI$$) –0.040 –0.018 –0.007 0.000 0.022 1,000 % industry 0.323 0.329 0.342 0.278 0.298 –0.024 % services 0.677 0.671 0.658 0.722 0.702 0.024 # employees 4.578 7.728 9.308 4.825 3.919 –0.007 % main province 0.236 0.238 0.222 0.173 0.144 –0.055 Debt per employee 128,460 175,598 164,244 165,833 114,852 –0.003 Utilized/granted credit 0.194 0.311 0.396 0.237 0.194 0.003 Multibanks 0.870 0.855 0.823 0.987 0.884 –0.017 NPLs 0.035 0.032 0.048 0.044 0.041 0.014 Quintile of exposure to credit supply shock Correlation with 1 2 3 4 5 credit supply ($$CSI$$) Credit supply index ($$CSI$$) –0.040 –0.018 –0.007 0.000 0.022 1,000 % industry 0.323 0.329 0.342 0.278 0.298 –0.024 % services 0.677 0.671 0.658 0.722 0.702 0.024 # employees 4.578 7.728 9.308 4.825 3.919 –0.007 % main province 0.236 0.238 0.222 0.173 0.144 –0.055 Debt per employee 128,460 175,598 164,244 165,833 114,852 –0.003 Utilized/granted credit 0.194 0.311 0.396 0.237 0.194 0.003 Multibanks 0.870 0.855 0.823 0.987 0.884 –0.017 NPLs 0.035 0.032 0.048 0.044 0.041 0.014 The table reports the average values of a set of firm-specific variables (by row) for each quintile of the sample distribution of the credit supply index ($$CSI$$). The % industry (services) is the share of firms in the industry (services) sector; the % main province is the percentage of firms that is located in the main province (i.e. Verona); Utilized/granted credit is the ratio between the utilized credit and total granted credit lines; Multibanks is a dummy equal to one if the firm has multiple banking relationship and equal to zero for firms borrowing from only one bank; and NPLs is a dummy equal to one if the firm has nonperforming loans at the beginning of the sample (December 2007). For the definition of $$CSI$$, see Section 2.2 and Equation (4). The last column reports the correlation between each of the row variables and the $$CSI$$ in the whole sample. Table 2 Orthogonality conditions Quintile of exposure to credit supply shock Correlation with 1 2 3 4 5 credit supply ($$CSI$$) Credit supply index ($$CSI$$) –0.040 –0.018 –0.007 0.000 0.022 1,000 % industry 0.323 0.329 0.342 0.278 0.298 –0.024 % services 0.677 0.671 0.658 0.722 0.702 0.024 # employees 4.578 7.728 9.308 4.825 3.919 –0.007 % main province 0.236 0.238 0.222 0.173 0.144 –0.055 Debt per employee 128,460 175,598 164,244 165,833 114,852 –0.003 Utilized/granted credit 0.194 0.311 0.396 0.237 0.194 0.003 Multibanks 0.870 0.855 0.823 0.987 0.884 –0.017 NPLs 0.035 0.032 0.048 0.044 0.041 0.014 Quintile of exposure to credit supply shock Correlation with 1 2 3 4 5 credit supply ($$CSI$$) Credit supply index ($$CSI$$) –0.040 –0.018 –0.007 0.000 0.022 1,000 % industry 0.323 0.329 0.342 0.278 0.298 –0.024 % services 0.677 0.671 0.658 0.722 0.702 0.024 # employees 4.578 7.728 9.308 4.825 3.919 –0.007 % main province 0.236 0.238 0.222 0.173 0.144 –0.055 Debt per employee 128,460 175,598 164,244 165,833 114,852 –0.003 Utilized/granted credit 0.194 0.311 0.396 0.237 0.194 0.003 Multibanks 0.870 0.855 0.823 0.987 0.884 –0.017 NPLs 0.035 0.032 0.048 0.044 0.041 0.014 The table reports the average values of a set of firm-specific variables (by row) for each quintile of the sample distribution of the credit supply index ($$CSI$$). The % industry (services) is the share of firms in the industry (services) sector; the % main province is the percentage of firms that is located in the main province (i.e. Verona); Utilized/granted credit is the ratio between the utilized credit and total granted credit lines; Multibanks is a dummy equal to one if the firm has multiple banking relationship and equal to zero for firms borrowing from only one bank; and NPLs is a dummy equal to one if the firm has nonperforming loans at the beginning of the sample (December 2007). For the definition of $$CSI$$, see Section 2.2 and Equation (4). The last column reports the correlation between each of the row variables and the $$CSI$$ in the whole sample. As far as the second term is concerned, bank-time fixed effects $$\hat{\delta}_{bt}$$ are exogenous by construction since they are purged of unobserved province-sector-quarter factors and it is rather implausible that unobserved effects at the firm level are able to affect nationwide banks’ lending policies. However, our identification assumption can be violated if banks with negative supply shocks were more likely to grant credit to firms that were hit more by the crisis. This may occur if, even in the same province-sector cluster, some banks can specialize into lending to firms with a specific demand for credit, since they rely on different product markets (i.e., more productive firms). In that case, the estimated bank-time fixed effects $$\hat{\delta}_{bt}$$ could capture a demand effect rather than a pure supply effect. Alternatively, it could be argued that there is an endogenous sorting between firms and banks, with weak banks lending to weak firms (Schivardi et al. 2017). In both cases we should observe some correlation between credit supply and firm characteristics. However, summary statistics reported in Table 2 shows that there is no systematic correlation between the size of the exposure to the credit supply shocks and a set of firm characteristics, such as size, financial dependence, banking relationships, leverage, bad credit history, geographical location, and sector of activity. The first five columns report summary statistics of firm beginning-of-the-period characteristics by quintile of $$CSI_{it}$$, averaged over the period 2008–2012, and the last column simplifies this information reporting the correlation between these pairs of variables. Firm characteristics are well balanced with respect to the average exposure to the credit shock during our temporal window. Moreover, for a subsample of firms for which we have balance sheet information, we can extend this exercise and show that the instrument is not correlated with labor productivity, leverage and riskiness, which could be taken as different proxies for firm quality (see the Online Table A12 and Appendix A-III).17 Our approach depart from Greenstone et al. (2014) along several dimensions that reinforce the exogeneity of the instrument.18 First, one may argue that banks differentiate their policies over the territory and that local lending policies are influenced by local economic conditions. To address this concern, we estimate Equation (1) dropping the Veneto provinces, so that we exclude the effects of demand and supply factors in this region from the calculation of bank-time fixed effects.19 Moreover, it is worth noting that according to lending survey pursued by the Bank of Italy, there is no evidence that banks applied different lending policies across the four Italian macro-regions (see Figure A2 in the Online Appendix). Second, we translate bank-time fixed effects at the firm rather than at the aggregate (i.e., county) level. This approach further reinforces the exogeneity of the instruments because while one may argue that unobservable shock in a county may affect (nationwide) lending policies of banks (especially when the local market is sufficiently large with respect to the national credit market of a certain bank), this is less plausible in case of unobservable shock at the firm level. Third, our data allows the estimation of time-varying bank fixed effects after having controlled for province-sector-time unobserved factors, while Greenstone et al. (2014) control only for counties-time unobserved factors. This means that we are able to account for bank-specific demand shocks that may occur whenever banks specialize, within the same provinces, in lending to different sectors that perform differently each other. Fourth, in Italy government interventions in favor of the banking system has been very limited, contrary to what has happened in other European countries and in the United States. This implies that bank lending policies were not affected by constraints imposed by the government as conditions to receive public support and, therefore, that our estimates are not affected by this potential source of bias. 3. Results 3.1 Main results To illustrate the impact of the credit supply, Figure 7 plots the employment patterns for firms classified in two groups, depending on whether they were exposed over the period 2008–2012 to tighter or easier lending policies (i.e., $$CSI$$ below or above the median). More specifically, the plotted values are the residuals (average of the two groups) of a regression of the logarithm of employees on firm and quarter fixed effects, so that the residuals are on average equal to zero and their time patterns show the dynamics of employment for the two groups. The two lines suggest that less favorable lending conditions are associated with a decrease in employment and with a divergent dynamic with respect to firms who experienced a better access to credit. The following regression tables statistically substantiate this visual evidence. Figure 7 View largeDownload slide Credit supply and employment dynamics The figure plots the averages of the residuals of a regression of the logarithm of employees on firm and quarter fixed effects. Averages are computed for groups of firms facing a more favorable (solid line) and less favorable (dashed line) credit supply conditions, defined as the average $$CSI$$ over 2008–2012 above or below the median, respectively. Figure 7 View largeDownload slide Credit supply and employment dynamics The figure plots the averages of the residuals of a regression of the logarithm of employees on firm and quarter fixed effects. Averages are computed for groups of firms facing a more favorable (solid line) and less favorable (dashed line) credit supply conditions, defined as the average $$CSI$$ over 2008–2012 above or below the median, respectively. Table 3 reports the 2SLS estimates of the baseline model for the whole sample of firms, including firm and quarter fixed effects (Column 1), and time-varying industry, class size, and province fixed effects (Columns 2 to 4). In line with most of the literature and to adopt a conservative approach, standard errors are clustered at that bank level.20 Table 3 Baseline regressions: IV estimates 1st stage Dep. var: $$\Delta Loan_{t,t-1}$$ $$CSI_{t,t-1}$$ 0.0786*** 0.0779*** 0.0757*** 0.0750*** (0.0178) (0.0167) (0.0165) (0.0171) $$R^2$$ 0.160 0.161 0.162 0.162 2nd stage Dep. var: $$\Delta Employment_t$$ $$\Delta Loan_{t,t-1}$$ 0.438*** 0.439*** 0.446*** 0.364*** (0.142) (0.140) (0.136) (0.111) Observations 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 194.7 191.5 180.9 169.5 Firm FEs Yes Yes Yes Yes Quarter FEs Yes – – – Industry $$\times$$ quarter FEs No Yes Yes Yes Size $$\times$$ quarter FEs No No Yes Yes Province $$\times$$ quarter FEs No No No Yes 1st stage Dep. var: $$\Delta Loan_{t,t-1}$$ $$CSI_{t,t-1}$$ 0.0786*** 0.0779*** 0.0757*** 0.0750*** (0.0178) (0.0167) (0.0165) (0.0171) $$R^2$$ 0.160 0.161 0.162 0.162 2nd stage Dep. var: $$\Delta Employment_t$$ $$\Delta Loan_{t,t-1}$$ 0.438*** 0.439*** 0.446*** 0.364*** (0.142) (0.140) (0.136) (0.111) Observations 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 194.7 191.5 180.9 169.5 Firm FEs Yes Yes Yes Yes Quarter FEs Yes – – – Industry $$\times$$ quarter FEs No Yes Yes Yes Size $$\times$$ quarter FEs No No Yes Yes Province $$\times$$ quarter FEs No No No Yes The table reports the regression results of the 2SLS estimation of Equation (1). The top panel shows the first-stage results, while the bottom panel reports the second-stage results. The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$; $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. Both variables are calculated like in Equation (2), so that they are bounded between $$-2$$ and $$+2$$. $$CSI$$ is the credit supply index, as defined in Section 2.2 and Equation (4). All four regressions are based on the full sample and they differ because of the set of time and borrower fixed effects that are included, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 3 Baseline regressions: IV estimates 1st stage Dep. var: $$\Delta Loan_{t,t-1}$$ $$CSI_{t,t-1}$$ 0.0786*** 0.0779*** 0.0757*** 0.0750*** (0.0178) (0.0167) (0.0165) (0.0171) $$R^2$$ 0.160 0.161 0.162 0.162 2nd stage Dep. var: $$\Delta Employment_t$$ $$\Delta Loan_{t,t-1}$$ 0.438*** 0.439*** 0.446*** 0.364*** (0.142) (0.140) (0.136) (0.111) Observations 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 194.7 191.5 180.9 169.5 Firm FEs Yes Yes Yes Yes Quarter FEs Yes – – – Industry $$\times$$ quarter FEs No Yes Yes Yes Size $$\times$$ quarter FEs No No Yes Yes Province $$\times$$ quarter FEs No No No Yes 1st stage Dep. var: $$\Delta Loan_{t,t-1}$$ $$CSI_{t,t-1}$$ 0.0786*** 0.0779*** 0.0757*** 0.0750*** (0.0178) (0.0167) (0.0165) (0.0171) $$R^2$$ 0.160 0.161 0.162 0.162 2nd stage Dep. var: $$\Delta Employment_t$$ $$\Delta Loan_{t,t-1}$$ 0.438*** 0.439*** 0.446*** 0.364*** (0.142) (0.140) (0.136) (0.111) Observations 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 194.7 191.5 180.9 169.5 Firm FEs Yes Yes Yes Yes Quarter FEs Yes – – – Industry $$\times$$ quarter FEs No Yes Yes Yes Size $$\times$$ quarter FEs No No Yes Yes Province $$\times$$ quarter FEs No No No Yes The table reports the regression results of the 2SLS estimation of Equation (1). The top panel shows the first-stage results, while the bottom panel reports the second-stage results. The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$; $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. Both variables are calculated like in Equation (2), so that they are bounded between $$-2$$ and $$+2$$. $$CSI$$ is the credit supply index, as defined in Section 2.2 and Equation (4). All four regressions are based on the full sample and they differ because of the set of time and borrower fixed effects that are included, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. The top panel reports the first-stage estimates, which show that, as expected, the $$CSI$$ is positively associated with the change in used loans and the coefficients is precisely estimated. The relevance of the instrument is further confirmed by the value of the first-stage F-statistic, which ranges between 170 and 195, well above the critical value of 10 suggested by Staiger and Stock (1997) to avoid the weak instrument bias. The second-stage results—reported in the bottom panel—confirm the existing evidence about the negative effect of a credit supply shock on employment (Chodorow-Reich 2014; Bentolila, Jensen, and Jimenez 2017), since the change in used loans has a significant and economically large effect on the variation in employment at the firm level. Comparing the four different specifications shows that adding fixed effects reduces the employment effect of the credit crunch, as fixed effects capture time-varying borrower-specific demand and productivity shocks. In particular, the point estimate of the coefficient on $$\Delta Loan$$ are broadly stable around $$0.44$$ in Columns 1 to 3, when adding time-varying industry and class size fixed effects, but decreases to $$0.36$$ when time-varying industry, size, and province fixed effects are jointly added in the model (Column 4). This result is robust to the inclusion of further controls to absorb demand and productivity shocks, and to alternative definitions of our key variables (see Online Appendix A-III for details). From now on, we will take the specification of Column 4 as our baseline. The point estimate of the bank lending channel is $$0.36$$, meaning that a 10% contraction in bank lending over two quarters translates into a 3.6% reduction in employment.21 In relative terms, one standard deviation of the predicted change of used loan explains 18% of the standard deviation of employment. 3.2 Job contract heterogeneity As a main contribution of our analysis, we look at within-firm dynamics and zoom in on the composition of the labor force adjustment, to assess in which way firms changed their workforce. Given that we cannot reconstruct the stock of workers by type of contracts and by worker characteristics for all firms, we estimate Equation (1) taking the quarterly change of employment at the firm level for a given job or worker characteristic, scaled by the average stock of all firm’s workers over the quarter, as dependent variables.22 Therefore, differently from the baseline model, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the job contracts/workers. Lacking that information in our sample, we use the aggregate shares at the regional level, as compiled by from the National Institute of Statistics (“Labour Force Survey”), in order to provide an economic interpretation of our findings (see Table 1). At first, we consider open-ended and temporary contracts—which include fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers—to test whether firms react to more binding financing constraints by reducing the use of temporary contracts more than open-ended ones (Table 4, top-left panel). We find that the employment adjustment happens primarily through variation of temporary contracts, consistent with the idea that firms use mostly fixed-term workers to absorb employment volatility (Caggese and Cuñat 2008) and with lower termination costs for temporary contracts.23 The coefficient on $$\Delta Loan$$ is positive and statistically significant for both type of contracts, even though there is an overrepresentation of temporary workers among dismissed employees, as also discussed by Bentolila, Jensen, and Jimenez (2017) for Spain. Although temporary contracts account for only slightly more than one-tenth of total contracts in the workforce (Table 1), they bear more than half of the effect of the change in credit supply ($$0.191/0.364 = 0.52$$, where $$0.364$$ is the estimated coefficient of credit supply variation for the entire workforce) (see Table 3, Column 4). By contrast, open-ended contracts account for 89% of the workforce, but contributed to less than half (48%) of the change in employment due to the credit crunch.24 Table 4 Job contract heterogeneity Dep. var.: $$\Delta Employment_t$$ Contracts Reason for exit Open-ended Temporary Dismissal Expiry Quit $$\Delta Loan_{t,t-1}$$ 0.170*** 0.191** –0.0357 –0.159** –0.0334 (0.0612) (0.0873) (0.0242) (0.0765) (0.0320) Flows Transitions Inflows Outflows Fixed to open Full- to part-time Part- to full-time $$\Delta Loan_{t,t-1}$$ 0.0836 –0.277** 0.0219* –0.00155 –0.00414 (0.0542) (0.111) (0.0117) (0.00631) (0.00714) Observations 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 $$1^{st}$$-stage F-statistic 169.5 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Contracts Reason for exit Open-ended Temporary Dismissal Expiry Quit $$\Delta Loan_{t,t-1}$$ 0.170*** 0.191** –0.0357 –0.159** –0.0334 (0.0612) (0.0873) (0.0242) (0.0765) (0.0320) Flows Transitions Inflows Outflows Fixed to open Full- to part-time Part- to full-time $$\Delta Loan_{t,t-1}$$ 0.0836 –0.277** 0.0219* –0.00155 –0.00414 (0.0542) (0.111) (0.0117) (0.00631) (0.00714) Observations 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 $$1^{st}$$-stage F-statistic 169.5 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes The table reports the (second-stage) regression results of the 2SLS estimation of Equation (1). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of job contracts, as labeled in each column, divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the job contracts (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the credit supply index $$CSI$$, as defined in Section 2.2 and Equation (4). The top panel reports the results for two subsamples of open-ended and temporary contracts, and the three subsamples of contract termination (outflows) due to dismissal, expiration of the contract, or voluntary quit. Temporary contracts include fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. The bottom panel reports the results for the subsamples of changes in employment due to inflows or outflows, and the ones based on three different transitions: from temporary to open-ended contracts, from full-time to part-time jobs, and from part-time to full-time jobs. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 4 Job contract heterogeneity Dep. var.: $$\Delta Employment_t$$ Contracts Reason for exit Open-ended Temporary Dismissal Expiry Quit $$\Delta Loan_{t,t-1}$$ 0.170*** 0.191** –0.0357 –0.159** –0.0334 (0.0612) (0.0873) (0.0242) (0.0765) (0.0320) Flows Transitions Inflows Outflows Fixed to open Full- to part-time Part- to full-time $$\Delta Loan_{t,t-1}$$ 0.0836 –0.277** 0.0219* –0.00155 –0.00414 (0.0542) (0.111) (0.0117) (0.00631) (0.00714) Observations 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 $$1^{st}$$-stage F-statistic 169.5 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Contracts Reason for exit Open-ended Temporary Dismissal Expiry Quit $$\Delta Loan_{t,t-1}$$ 0.170*** 0.191** –0.0357 –0.159** –0.0334 (0.0612) (0.0873) (0.0242) (0.0765) (0.0320) Flows Transitions Inflows Outflows Fixed to open Full- to part-time Part- to full-time $$\Delta Loan_{t,t-1}$$ 0.0836 –0.277** 0.0219* –0.00155 –0.00414 (0.0542) (0.111) (0.0117) (0.00631) (0.00714) Observations 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 $$1^{st}$$-stage F-statistic 169.5 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes The table reports the (second-stage) regression results of the 2SLS estimation of Equation (1). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of job contracts, as labeled in each column, divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the job contracts (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the credit supply index $$CSI$$, as defined in Section 2.2 and Equation (4). The top panel reports the results for two subsamples of open-ended and temporary contracts, and the three subsamples of contract termination (outflows) due to dismissal, expiration of the contract, or voluntary quit. Temporary contracts include fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. The bottom panel reports the results for the subsamples of changes in employment due to inflows or outflows, and the ones based on three different transitions: from temporary to open-ended contracts, from full-time to part-time jobs, and from part-time to full-time jobs. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. To better understand the employment dynamics following the credit crunch, we differentiate between inflows and outflows and we find that our results are mostly driven by the dynamics of outflows, which are higher for firms more exposed to the credit supply shock, even though the effect on inflows is also marginally significant (Table 4, bottom-left panel). Then, within outflows, we differentiate across the possible reasons of the exit and we find evidence that outflows are exclusively due to nonrenewal of expired contracts, while there is no evidence that the adjustment works through dismissal or quit (Table 4, top-right panel). Finally, we look at the transitions across job contracts, considering both contract type and time schedule. We find evidence that firms more exposed to negative credit shocks are less likely to transform temporary contracts into open-ended ones, while it seems that financing constraints do not affect firm policies in terms of transition between part-time and full-time jobs (Table 4, bottom-right panel). 3.3 Worker heterogeneity An alternative interpretation of the fact that the adjustment is mainly borne by temporary contracts could be related to the presence of a high EPL, which makes firing permanent workers for Italian firms very difficult. If terminating open-ended contracts is indeed difficult and costly, the concentration of the adjustment on temporary contracts does not come as a surprise. To mitigate this concern, in this section we take advantage of additional dimensions in which we can slice our data to measure the impact of the credit crunch on employment, conditional both on contract type and on a number of workers’ characteristics.25 We first differentiate across three levels of education: low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education), based on the ISCED classification. We observe that firms that have experienced a reduction in the supply of credit reacted reducing mostly the employment of low- and medium-educated workers, while the effect for the high-educated ones is smaller and only marginally significant (Table 5, top panel). In particular, using the relative shares reported in Table 1, the elasticity of employment to credit supply for low-educated workers is higher than the average and equal to $$0.48$$ ($$= 0.186/0.387$$). The corresponding elasticities are equal to $$0.31$$ and $$0.19$$ for medium- and high-educated workers, respectively. In other words, changes in employment within low-educated workers account for more than half of the total effect of $$\Delta Loan$$ ($$0.186/0.364 = 0.51$$), even though low-educated workers account for less than 40% of the workforce.26 Table 5 Worker heterogeneity by education and contract type Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.364*** 0.186*** 0.151*** 0.0261* (0.111) (0.0393) (0.0366) (0.0134) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.170*** 0.0684** 0.0817*** 0.0215** (0.0612) (0.0299) (0.0308) (0.00862) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.191** 0.115** 0.0717** 0.00421 (0.0873) (0.0550) (0.0363) (0.0119) Observations 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.364*** 0.186*** 0.151*** 0.0261* (0.111) (0.0393) (0.0366) (0.0134) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.170*** 0.0684** 0.0817*** 0.0215** (0.0612) (0.0299) (0.0308) (0.00862) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.191** 0.115** 0.0717** 0.00421 (0.0873) (0.0550) (0.0363) (0.0119) Observations 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes The table reports the (second-stage) regression results of the 2SLS estimation of Equation (1). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the credit supply index $$CSI$$, as defined in Section 2.2 and Equation (4). The table reports the results for all workers, regardless of their education level (first column), and for those with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education based on the ISCED classification. Results reported in the top panel refer all contracts, while those in the middle panel to open-ended contracts, and the ones in the bottom panel refer to temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 5 Worker heterogeneity by education and contract type Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.364*** 0.186*** 0.151*** 0.0261* (0.111) (0.0393) (0.0366) (0.0134) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.170*** 0.0684** 0.0817*** 0.0215** (0.0612) (0.0299) (0.0308) (0.00862) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.191** 0.115** 0.0717** 0.00421 (0.0873) (0.0550) (0.0363) (0.0119) Observations 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.364*** 0.186*** 0.151*** 0.0261* (0.111) (0.0393) (0.0366) (0.0134) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.170*** 0.0684** 0.0817*** 0.0215** (0.0612) (0.0299) (0.0308) (0.00862) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.191** 0.115** 0.0717** 0.00421 (0.0873) (0.0550) (0.0363) (0.0119) Observations 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes The table reports the (second-stage) regression results of the 2SLS estimation of Equation (1). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the credit supply index $$CSI$$, as defined in Section 2.2 and Equation (4). The table reports the results for all workers, regardless of their education level (first column), and for those with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education based on the ISCED classification. Results reported in the top panel refer all contracts, while those in the middle panel to open-ended contracts, and the ones in the bottom panel refer to temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Then we combine the effect of contract type with worker education. Results are reported in the middle and bottom panels of Table 5, and visualized in Figure 8, which shows in the dark bars the contribution of the overall estimated effect of credit supply on employment ($$0.364$$) due to the combination of contract type and education levels. For comparison, the white bars report the share in total employment by contract type and education. In all relevant cases discussed below, the difference between the two bars is statistically significant. We find that firms adjusted their labor force in response to a contraction in the supply of credit predominantly reducing temporary contracts of low- and medium-educated workers, even though they account for a relatively low share of total employment. By contrast, high-educated workers have been able to insulate themselves, even if hired with temporary contracts. The effect on low-educated workers with a temporary contract accounts for 32% of the total employment effect ($$0.115/0.364 = 0.32$$), even though they represent less than 4% of the workforce. This share declines to 19% moving to an open-ended contract (but they account for 33% of the workforce) and further down to 6% for a high-educated worker with an open-ended contract, which account for 11% of the workforce (the effect is not significant for high-educated workers with temporary contracts). These results are consistent with the hypothesis that low-skilled individuals suffer most from recessions, as firms follow a skill upgrading strategy (Reder 1955; Hershbein and Kahn 2016), and with the empirical evidence on Germany discussed by Hochfellner et al. (2016). Overall, our results indicate that the combination of low-education and temporary contract identifies the profile of workers who have been hit by the credit crunch, while high education makes the difference between temporary and open-ended contracts almost irrelevant. Figure 8 View largeDownload slide The effect of the credit crunch by contract type and education The dark bars plot the contribution of the overall estimated effect of the credit supply on employment ($$0.364$$) for the combination of contract type (temporary contracts and open-ended contracts) and education levels (low, medium, and high). These relative shares are based on the estimates reported in Table 5. The white bars report the share in total employment by contract type and education, as reported in Table 1. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. Workers are divided across low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education levels based on the ISCED classification. Figure 8 View largeDownload slide The effect of the credit crunch by contract type and education The dark bars plot the contribution of the overall estimated effect of the credit supply on employment ($$0.364$$) for the combination of contract type (temporary contracts and open-ended contracts) and education levels (low, medium, and high). These relative shares are based on the estimates reported in Table 5. The white bars report the share in total employment by contract type and education, as reported in Table 1. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. Workers are divided across low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education levels based on the ISCED classification. Second, we assess whether firms adjusted their labor force differentiating across workers, depending on their gender, age, and nationality. As before, we first look at the whole sample (Table 6, top panel) and we then differentiate between contract type (middle and bottom panels). Our results indicate that the employment effect in response to a reduction in the supply of credit is concentrated among women, foreign and younger workers. In particular, female workers represent around 40% of total employment, but they account for a significantly larger share—60% ($$0.220/0.364$$)—of the total change in employment. Similarly, foreign workers are less the 10% of the labor force, but their employment dynamics explains more than 24% of the total change in employment.27 Results suggest that younger people are more likely to feel the consequences of the credit crunch, consistent with recent evidence showing that young workers are the most affected during recessions (Forsythe 2016). Workers under 30 contribute to slightly less than one-third of the overall employment effect, a value statistically larger than their share in the workforce, equal to 18%. Table 6 Worker heterogeneity by personal characteristics and contract type Dep. var.: $$\Delta Employment_t$$ Gender Age Nationality Male Female Under 30 Over 30 Italian Foreign All contracts $$\Delta Loan_{t,t-1}$$ 0.144** 0.220*** 0.105*** 0.258*** 0.274*** 0.0885** (0.0573) (0.0686) (0.0402) (0.0807) (0.0850) (0.0353) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.0785* 0.0941*** 0.0355*** 0.138*** 0.130*** 0.0423** (0.0411) (0.0256) (0.0130) (0.0525) (0.0501) (0.0168) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.0610* 0.131** 0.0737* 0.116** 0.140** 0.0477 (0.0350) (0.0610) (0.0396) (0.0554) (0.0613) (0.0308) Observations 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 169.5 169.5 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Gender Age Nationality Male Female Under 30 Over 30 Italian Foreign All contracts $$\Delta Loan_{t,t-1}$$ 0.144** 0.220*** 0.105*** 0.258*** 0.274*** 0.0885** (0.0573) (0.0686) (0.0402) (0.0807) (0.0850) (0.0353) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.0785* 0.0941*** 0.0355*** 0.138*** 0.130*** 0.0423** (0.0411) (0.0256) (0.0130) (0.0525) (0.0501) (0.0168) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.0610* 0.131** 0.0737* 0.116** 0.140** 0.0477 (0.0350) (0.0610) (0.0396) (0.0554) (0.0613) (0.0308) Observations 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 169.5 169.5 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes The table reports the (second-stage) regression results of the 2SLS estimation of Equation (1). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of workers, as labeled in each column, divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the credit supply index $$CSI$$, as defined in Section 2.2 and Equation (4). The left panel reports the results for the subsamples of men and women. The middle panel reports the results for the subsamples of workers whose age is below or above 30 years. The right panel show the results for the subsample of Italian and foreign workers. Results reported in the top panel refer all contracts, while those in the middle panel to open-ended contracts, and the ones in the bottom panel refer to temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 6 Worker heterogeneity by personal characteristics and contract type Dep. var.: $$\Delta Employment_t$$ Gender Age Nationality Male Female Under 30 Over 30 Italian Foreign All contracts $$\Delta Loan_{t,t-1}$$ 0.144** 0.220*** 0.105*** 0.258*** 0.274*** 0.0885** (0.0573) (0.0686) (0.0402) (0.0807) (0.0850) (0.0353) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.0785* 0.0941*** 0.0355*** 0.138*** 0.130*** 0.0423** (0.0411) (0.0256) (0.0130) (0.0525) (0.0501) (0.0168) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.0610* 0.131** 0.0737* 0.116** 0.140** 0.0477 (0.0350) (0.0610) (0.0396) (0.0554) (0.0613) (0.0308) Observations 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 169.5 169.5 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Gender Age Nationality Male Female Under 30 Over 30 Italian Foreign All contracts $$\Delta Loan_{t,t-1}$$ 0.144** 0.220*** 0.105*** 0.258*** 0.274*** 0.0885** (0.0573) (0.0686) (0.0402) (0.0807) (0.0850) (0.0353) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.0785* 0.0941*** 0.0355*** 0.138*** 0.130*** 0.0423** (0.0411) (0.0256) (0.0130) (0.0525) (0.0501) (0.0168) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.0610* 0.131** 0.0737* 0.116** 0.140** 0.0477 (0.0350) (0.0610) (0.0396) (0.0554) (0.0613) (0.0308) Observations 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 2,459,949 1st-stage F-statistic 169.5 169.5 169.5 169.5 169.5 169.5 Firm FEs Yes Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes Yes The table reports the (second-stage) regression results of the 2SLS estimation of Equation (1). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of workers, as labeled in each column, divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the credit supply index $$CSI$$, as defined in Section 2.2 and Equation (4). The left panel reports the results for the subsamples of men and women. The middle panel reports the results for the subsamples of workers whose age is below or above 30 years. The right panel show the results for the subsample of Italian and foreign workers. Results reported in the top panel refer all contracts, while those in the middle panel to open-ended contracts, and the ones in the bottom panel refer to temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. These findings could reflect a propensity of women, foreign and younger workers to have temporary contracts, so that they naturally end up being the most affected if firms primarily adjust cutting back on temporary rather than open-ended contracts. To address this concern we replicate what done in Table 5 and we consider separately temporary and open-ended contracts. Results indicate that both women and foreign workers are more likely to be affected by the credit crunch regardless of the kind of contract, while the overall effect found for young workers is driven exclusively by those employed with temporary contracts.28 Overall, the fact that less educated workers, as well as women and foreign workers, are relatively more likely to lose their job, even within workers employed with temporary contracts, suggests that EPL is not the only driving force behind the adjustment in employment.29 Moreover, the significant concentration of the employment adjustment on less educated workers and low-skill occupations is consistent with a productivity-enhancing reallocation and with recent evidence showing a cleansing effect of the Great Recession (Foster et al. 2016). 4. Extensions 4.1 Adjustment at the extensive and intensive margins So far our analysis has considered the effects of a financial shock at the extensive and intensive margins together. However, understanding if the aggregate employment effect is driven by a downsizing of the workforce in active firms or by firm closures has important implications for the understanding the crisis and of the mechanisms of workforce management within the firm. To shed some light on the margins of adjustment, we first reestimate our model on a subsample that excludes the firms that close down in a given quarter. Specifically, in each quarter we consider all active firms that can adjust at the intensive margin and the ones that will close in future quarters, but that can still adjust their workforce in the quarters before closure. Then, to look at the extensive margin, we estimate a linear probability model for the likelihood that a firm closes its activity in a given quarter. Our results, reported in Table 7, indicate that the adjustment to a contraction in credit supply has happened both at the intensive and extensive margins, in line with the evidence on Spain (Bentolila, Jensen, and Jimenez 2017). When we drop from the sample firm closures, we still find a precisely identified elasticity, even though its magnitude is smaller, as a 10% contraction in credit translates into a 2.5% fall in employment. In addition, the adjustment at the intensive margin falls disproportionately on temporary workers, which account for about three quarters of the fall in employment (the effect is about 50% in the whole sample). Hence, part of the effect on open-ended contracts is due to firm exit, consistent with the presence of labor market rigidities and high dismissal costs for open-ended contracts. Finally, the last column shows that a shortfall in the supply of credit increases the likelihood of firm exit. This effect is economically meaningful: considering the average contraction of bank credit of 1.6% in the sample period, the estimated coefficient implies a 0.1% increase in the probability that a firm closes down, which accounts for about one seventh of the average exit rate (Table 1). Table 7 Adjustment at the intensive and extensive margins Dep. var.: $$\Delta Employment_t$$ $$EXIT_t$$ Full sample Excluding firm closures Full sample All contracts Open-ended Temporary (1) (2) (3) (4) (5) $$\Delta Loan_{t,t-1}$$ 0.364*** 0.253*** 0.0653** 0.185** –0.0591** (0.111) (0.0914) (0.0282) (0.0886) (0.0269) Observations 2,459,949 2,443,652 2,443,652 2,443,652 2,459,949 1st-stage F-statistic 169.5 169.1 169.1 169.1 169.5 Firm FEs Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ $$EXIT_t$$ Full sample Excluding firm closures Full sample All contracts Open-ended Temporary (1) (2) (3) (4) (5) $$\Delta Loan_{t,t-1}$$ 0.364*** 0.253*** 0.0653** 0.185** –0.0591** (0.111) (0.0914) (0.0282) (0.0886) (0.0269) Observations 2,459,949 2,443,652 2,443,652 2,443,652 2,459,949 1st-stage F-statistic 169.5 169.1 169.1 169.1 169.5 Firm FEs Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes The table reports the (second-stage) regression results of the 2SLS estimation of Equation (1). The dependent variable is: $$\Delta Employment_t$$, defined as the change in employment at the firm level over the year $$t$$ (Columns 1 and 4); and $$EXIT_t$$, defined as a dichotomous variable equal to one if the firm closed in the quarter $$t$$ but was still in operation in the previous quarter $$t-1$$, and zero elsewhere (c]Column 5). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in the year $$t-1$$. $$\Delta Employment_t$$ and $$\Delta Loan_{t,t-1}$$ are calculated like in Equation (2), so that they are bounded between $$-2$$ and $$+2$$. In the first-stage regressions, the excluded instrument is the credit supply index $$CSI$$, as defined in Section 2.2 and Equation (4). Results in Columns 1 and 5 are based on the full sample, while all other results are based on the subsample that excludes firm closures (i.e., a firm that closes in a given quarter is still in the sample for the previous quarters, when it was active). Results for this subsample are reported both for all job contracts (Column 2) and separated for the different types of contracts (open-ended and temporary, Columns 3 and 4). Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All linear regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 7 Adjustment at the intensive and extensive margins Dep. var.: $$\Delta Employment_t$$ $$EXIT_t$$ Full sample Excluding firm closures Full sample All contracts Open-ended Temporary (1) (2) (3) (4) (5) $$\Delta Loan_{t,t-1}$$ 0.364*** 0.253*** 0.0653** 0.185** –0.0591** (0.111) (0.0914) (0.0282) (0.0886) (0.0269) Observations 2,459,949 2,443,652 2,443,652 2,443,652 2,459,949 1st-stage F-statistic 169.5 169.1 169.1 169.1 169.5 Firm FEs Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ $$EXIT_t$$ Full sample Excluding firm closures Full sample All contracts Open-ended Temporary (1) (2) (3) (4) (5) $$\Delta Loan_{t,t-1}$$ 0.364*** 0.253*** 0.0653** 0.185** –0.0591** (0.111) (0.0914) (0.0282) (0.0886) (0.0269) Observations 2,459,949 2,443,652 2,443,652 2,443,652 2,459,949 1st-stage F-statistic 169.5 169.1 169.1 169.1 169.5 Firm FEs Yes Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Yes The table reports the (second-stage) regression results of the 2SLS estimation of Equation (1). The dependent variable is: $$\Delta Employment_t$$, defined as the change in employment at the firm level over the year $$t$$ (Columns 1 and 4); and $$EXIT_t$$, defined as a dichotomous variable equal to one if the firm closed in the quarter $$t$$ but was still in operation in the previous quarter $$t-1$$, and zero elsewhere (c]Column 5). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in the year $$t-1$$. $$\Delta Employment_t$$ and $$\Delta Loan_{t,t-1}$$ are calculated like in Equation (2), so that they are bounded between $$-2$$ and $$+2$$. In the first-stage regressions, the excluded instrument is the credit supply index $$CSI$$, as defined in Section 2.2 and Equation (4). Results in Columns 1 and 5 are based on the full sample, while all other results are based on the subsample that excludes firm closures (i.e., a firm that closes in a given quarter is still in the sample for the previous quarters, when it was active). Results for this subsample are reported both for all job contracts (Column 2) and separated for the different types of contracts (open-ended and temporary, Columns 3 and 4). Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All linear regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Given that the composition of the adjustment at the intensive margin looks different that the overall effect, we replicate the analysis discussed in Section 3.3 to look at the role of worker heterogeneity by education in the restricted sample that excludes firm closures. The results are qualitatively similar, but stronger than those obtained in the whole sample, suggesting that the reduction in employment due to firm exit has been relatively more homogeneous across contracts and workers than the one that involved active firms. In particular, firms which experienced a reduction in the supply of credit, but did not close, reduced employment mostly among low- and medium-educated workers (Table 8, top panel). Then considering contract type and education together clearly reinforces one of our main findings. The intensive margin adjustment has exclusively affected less educated workers with temporary contracts, whereas high-educated temporary workers—which represent about one-fifth of all temporary contracts— have not been hit by the financial shock (Table 8, middle and bottom panels). Table 8 The effect of contract type and education: Intensive margin Dep. var.: $$\Delta Employment_t$$ Education level Low Medium High All contract $$\Delta Loan_{t,t-1}$$ 0.141** 0.101*** 0.0108 (0.0604) (0.0387) (0.0121) Open-ended contract $$\Delta Loan_{t,t-1}$$ 0.0261 0.0307** 0.00897 (0.0219) (0.0152) (0.00637) Temporary contract $$\Delta Loan_{t,t-1}$$ 0.114** 0.0695* 0.00181 (0.0562) (0.0370) (0.0115) Observations 2,443,652 2,443,652 2,443,652 1st-stage F-statistic 169.1 169.1 169.1 Firm FEs Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level Low Medium High All contract $$\Delta Loan_{t,t-1}$$ 0.141** 0.101*** 0.0108 (0.0604) (0.0387) (0.0121) Open-ended contract $$\Delta Loan_{t,t-1}$$ 0.0261 0.0307** 0.00897 (0.0219) (0.0152) (0.00637) Temporary contract $$\Delta Loan_{t,t-1}$$ 0.114** 0.0695* 0.00181 (0.0562) (0.0370) (0.0115) Observations 2,443,652 2,443,652 2,443,652 1st-stage F-statistic 169.1 169.1 169.1 Firm FEs Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes The table reports the (second-stage) regression results of the 2SLS estimation of Equation (1) on a restricted sample that excludes firm closures (i.e. a firm that closes in a given quarter is still in the sample for the previous quarters, when it was active). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the credit supply index $$CSI$$, as defined in Section 2.2 and Equation (4). Results reported in the top panel refer to all job contracts, the ones reported in the middle panel to open-ended contracts, and the ones reported in the bottom panel refer to fixed-ended contracts, for different level of worker education. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. Education level are based on the ISCED classification: low means at most compulsory education, medium is at most upper secondary education, and high indicates tertiary education. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 8 The effect of contract type and education: Intensive margin Dep. var.: $$\Delta Employment_t$$ Education level Low Medium High All contract $$\Delta Loan_{t,t-1}$$ 0.141** 0.101*** 0.0108 (0.0604) (0.0387) (0.0121) Open-ended contract $$\Delta Loan_{t,t-1}$$ 0.0261 0.0307** 0.00897 (0.0219) (0.0152) (0.00637) Temporary contract $$\Delta Loan_{t,t-1}$$ 0.114** 0.0695* 0.00181 (0.0562) (0.0370) (0.0115) Observations 2,443,652 2,443,652 2,443,652 1st-stage F-statistic 169.1 169.1 169.1 Firm FEs Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level Low Medium High All contract $$\Delta Loan_{t,t-1}$$ 0.141** 0.101*** 0.0108 (0.0604) (0.0387) (0.0121) Open-ended contract $$\Delta Loan_{t,t-1}$$ 0.0261 0.0307** 0.00897 (0.0219) (0.0152) (0.00637) Temporary contract $$\Delta Loan_{t,t-1}$$ 0.114** 0.0695* 0.00181 (0.0562) (0.0370) (0.0115) Observations 2,443,652 2,443,652 2,443,652 1st-stage F-statistic 169.1 169.1 169.1 Firm FEs Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes The table reports the (second-stage) regression results of the 2SLS estimation of Equation (1) on a restricted sample that excludes firm closures (i.e. a firm that closes in a given quarter is still in the sample for the previous quarters, when it was active). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the credit supply index $$CSI$$, as defined in Section 2.2 and Equation (4). Results reported in the top panel refer to all job contracts, the ones reported in the middle panel to open-ended contracts, and the ones reported in the bottom panel refer to fixed-ended contracts, for different level of worker education. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. Education level are based on the ISCED classification: low means at most compulsory education, medium is at most upper secondary education, and high indicates tertiary education. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. 4.2 General equilibrium effects Overall, results indicate that the effect of the credit crunch on employment is economically relevant. Our findings are roughly comparable, in magnitude, to those estimated by Bentolila, Jensen, and Jimenez (2017) for Spain and Chodorow-Reich (2014) for the United States. However, compared to these exercises—which are generally focused on medium and large enterprises—our analysis is less subject to external validity concerns related to the representativeness of the data, since our sample include micro and small firms and covers almost the universe of private nonfinancial firms and employment of the region.30 However, the contraction in employment estimated at the firm level could be offset by the behavior of firms that are not included in our analysis because they do not rely on bank credit (i.e. more formally, all firms which do not have a match in the credit register); for instance, temporary workers—which are more likely to lose their job in our sample—could be hired by nonbanked firms. In that case, the welfare implications of our analysis could differ. To estimate the general equilibrium effects of the credit contraction, we run a set of regressions at the province-industry-quarter level, where we aggregate employment and loan outcomes considering all firms in the region, including the ones without a match in the credit register. The credit supply index is also computed aggregating bank-specific CSIs in each province-industry-quarter cluster. Results, reported in Table 9, show that the elasticity of aggregate employment to bank lending remains economically relevant, suggesting that any offsetting effect of the credit crunch due to firms without banking relationships, if present, has been relatively small. Moreover, considering separately temporary and open-ended contracts and workers with different education levels shows that the effect is again mostly driven by less educated workers and those with temporary contracts. Table 9 General equilibrium effects Dep. var.: $$\Delta Employment_t$$ Whole Contracts Education level sample Open-ended Temporary Low Medium High $$\Delta Loan_{t,t-1}$$ 0.408** 0.140** 0.265* 0.140* 0.218** 0.0540 (0.168) (0.0618) (0.135) (0.0775) (0.0911) (0.0380) Observations 3,780 3,780 3,780 3,780 3,780 3,780 1st-stage F-statistic 14.52 14.52 14.52 14.52 14.52 14.52 Industry FEs Yes Yes Yes Yes Yes Yes Province FEs Yes Yes Yes Yes Yes Yes Quarter FEs Yes Yes Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Whole Contracts Education level sample Open-ended Temporary Low Medium High $$\Delta Loan_{t,t-1}$$ 0.408** 0.140** 0.265* 0.140* 0.218** 0.0540 (0.168) (0.0618) (0.135) (0.0775) (0.0911) (0.0380) Observations 3,780 3,780 3,780 3,780 3,780 3,780 1st-stage F-statistic 14.52 14.52 14.52 14.52 14.52 14.52 Industry FEs Yes Yes Yes Yes Yes Yes Province FEs Yes Yes Yes Yes Yes Yes Quarter FEs Yes Yes Yes Yes Yes Yes The table reports the regression results of the 2SLS estimation of the analogous of Equation (1) on data aggregated at the province-industry-sector level, considering all firms in the region. The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the province-industry-sector level over the quarter $$t$$; $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the province-industry-sector in quarters $$t$$ and $$t-1$$. Both variables are calculated like in Equation (2), so that they are bounded between $$-2$$ and $$+2$$. $$CSI$$ is the credit supply index, as defined in Section 2.2 and Equation (4) and calculated aggregating bank-specific CSIs at the province-industry-sector level. The first column reports the results for the whole sample, the others report the results for two subsamples of open-ended and temporary contracts, and the three subsamples of workers with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education based on the ISCED classification. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions include a set of time, industry and province fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries and 7 provinces. Robust standard errors are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 9 General equilibrium effects Dep. var.: $$\Delta Employment_t$$ Whole Contracts Education level sample Open-ended Temporary Low Medium High $$\Delta Loan_{t,t-1}$$ 0.408** 0.140** 0.265* 0.140* 0.218** 0.0540 (0.168) (0.0618) (0.135) (0.0775) (0.0911) (0.0380) Observations 3,780 3,780 3,780 3,780 3,780 3,780 1st-stage F-statistic 14.52 14.52 14.52 14.52 14.52 14.52 Industry FEs Yes Yes Yes Yes Yes Yes Province FEs Yes Yes Yes Yes Yes Yes Quarter FEs Yes Yes Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Whole Contracts Education level sample Open-ended Temporary Low Medium High $$\Delta Loan_{t,t-1}$$ 0.408** 0.140** 0.265* 0.140* 0.218** 0.0540 (0.168) (0.0618) (0.135) (0.0775) (0.0911) (0.0380) Observations 3,780 3,780 3,780 3,780 3,780 3,780 1st-stage F-statistic 14.52 14.52 14.52 14.52 14.52 14.52 Industry FEs Yes Yes Yes Yes Yes Yes Province FEs Yes Yes Yes Yes Yes Yes Quarter FEs Yes Yes Yes Yes Yes Yes The table reports the regression results of the 2SLS estimation of the analogous of Equation (1) on data aggregated at the province-industry-sector level, considering all firms in the region. The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the province-industry-sector level over the quarter $$t$$; $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the province-industry-sector in quarters $$t$$ and $$t-1$$. Both variables are calculated like in Equation (2), so that they are bounded between $$-2$$ and $$+2$$. $$CSI$$ is the credit supply index, as defined in Section 2.2 and Equation (4) and calculated aggregating bank-specific CSIs at the province-industry-sector level. The first column reports the results for the whole sample, the others report the results for two subsamples of open-ended and temporary contracts, and the three subsamples of workers with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education based on the ISCED classification. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions include a set of time, industry and province fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries and 7 provinces. Robust standard errors are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. 5. Robustness 5.1 Internal validity: Credit supply index at the firm level In Section 2.2 we have discussed the exogeneity of our measure of credit supply and provided some empirical evidence to support this assumption. In the end, the key identifying assumption is that firm-level loan demand is not bank-specific and, even though the orthogonality conditions show that the $$CSI$$ is uncorrelated with a large number of observable characteristics, we cannot test the excluding restriction that changes in credit demand are not correlated with the credit supply index. Thus, to further strengthen the internal validity of our results, we construct the $$CSI$$ based on a modified version of Equation (3), which exploits the subsample of firms borrowing from multiple banks (Khwaja and Mian 2008): \begin{align}\label{eq:bank2} \Delta L_{bit} = \alpha + \delta_{bt} + \gamma_{it} + \epsilon_{bit}, \end{align} (5) where the outcome variable $$\Delta L_{bit}$$ is the percentage change in outstanding business loans by bank $$b$$ to firm $$i$$ at time $$t$$. In this case, the $$\gamma_{it}$$ fixed effects absorb firm-specific time-varying credit demand, rather than assuming that all firms in the same province-sector cluster have the same demand. However, this choice comes at the cost of identifying the nationwide bank lending policies parameters $$\delta_{bt}$$ on the subsample of firms with multiple bank relationships. As those firms are likely to be different from those with only one bank under a number of characteristics (in our data, e.g., more than 90% of medium and large firms have multiple relationships in contrast to about 30% for micro-firms), the identification of $$\delta_{bt}$$ could be affected by the difference in the composition of the sample. Results are shown in Table 10 and they are consistent with our baseline regressions. The average effect of the contraction in credit supply on employment is driven by temporary contracts and by less educated workers. However, the estimated elasticity is smaller than that estimated with the $$CSI$$ constructed at the province-sector level. A possible explanation of the smaller magnitude of the coefficient on $$\Delta Loan$$ could be due to the sample used to estimate the bank lending policies $$\delta_{bt}$$: excluding firms with single bank relationships—which are overrepresented among small firms borrowing from small banks—could imply a limited capacity to account for credit demand, leading to a weaker link between credit supply and bank lending policies. Table 10 Credit supply index at the firm level Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.145** 0.0841** 0.0649** –0.000487 (0.0581) (0.0335) (0.0318) (0.0133) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.0154 0.00716 0.00970 –4.61e-07 (0.0405) (0.0221) (0.0219) (0.00793) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.131*** 0.0772*** 0.0547** –0.000517 (0.0395) (0.0251) (0.0236) (0.0109) Observations 2,459,785 2,459,786 2,459,787 2,459,788 1st-stage F-statistic 202.2 202.3 202.4 202.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.145** 0.0841** 0.0649** –0.000487 (0.0581) (0.0335) (0.0318) (0.0133) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.0154 0.00716 0.00970 –4.61e-07 (0.0405) (0.0221) (0.0219) (0.00793) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.131*** 0.0772*** 0.0547** –0.000517 (0.0395) (0.0251) (0.0236) (0.0109) Observations 2,459,785 2,459,786 2,459,787 2,459,788 1st-stage F-statistic 202.2 202.3 202.4 202.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes The table reports the regression results of the 2SLS estimation of Equation (1) on the whole sample. The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the alternative credit supply index $$CSI$$, built from a nationwide bank lending regression that include firm$$\times$$quarter fixed effects (see Equation (5) and Section 5.1). The table reports the results for all workers, regardless of their education level (first column), and for those with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education based on the ISCED classification. Results reported in the top panel refer all contracts, while those in the middle panel to open-ended contracts, and the ones in the bottom panel refer to temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the firm level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 10 Credit supply index at the firm level Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.145** 0.0841** 0.0649** –0.000487 (0.0581) (0.0335) (0.0318) (0.0133) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.0154 0.00716 0.00970 –4.61e-07 (0.0405) (0.0221) (0.0219) (0.00793) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.131*** 0.0772*** 0.0547** –0.000517 (0.0395) (0.0251) (0.0236) (0.0109) Observations 2,459,785 2,459,786 2,459,787 2,459,788 1st-stage F-statistic 202.2 202.3 202.4 202.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.145** 0.0841** 0.0649** –0.000487 (0.0581) (0.0335) (0.0318) (0.0133) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.0154 0.00716 0.00970 –4.61e-07 (0.0405) (0.0221) (0.0219) (0.00793) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.131*** 0.0772*** 0.0547** –0.000517 (0.0395) (0.0251) (0.0236) (0.0109) Observations 2,459,785 2,459,786 2,459,787 2,459,788 1st-stage F-statistic 202.2 202.3 202.4 202.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes The table reports the regression results of the 2SLS estimation of Equation (1) on the whole sample. The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the alternative credit supply index $$CSI$$, built from a nationwide bank lending regression that include firm$$\times$$quarter fixed effects (see Equation (5) and Section 5.1). The table reports the results for all workers, regardless of their education level (first column), and for those with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education based on the ISCED classification. Results reported in the top panel refer all contracts, while those in the middle panel to open-ended contracts, and the ones in the bottom panel refer to temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the firm level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. 5.2 External validity: The employment effect of the credit crunch in crises times In Section 1 we have provided a broad set of statistics to support the fact that the Veneto region could be considered a representative case study. Another issue that may affect the external validity of our results is the specific time period under analysis, which covers the Lehman’s bankruptcy and the European sovereign debt crisis.31 As job contracts data are not available before 2008, we are not able to do a standard comparison between a crisis and a tranquil period. However, to take into account the significant changes in economic and financial conditions over the four years of our sample, we split the analysis in 2 subsamples: the first one covers the Lehman crisis and ends in 2011:Q2, whereas the second one excludes the Lehman shock and, starting in 2009:Q2, focuses instead on the European sovereign debt crisis. Separating the two episodes is of interest since the dynamics of the banking crisis in Italy has changed over time. The crisis has been mostly concentrated among a few large banks at the beginning and became more widespread in coincidence with the sovereign debt crisis, when the tightening of credit conditions reflected the common shock of widening sovereign spreads, rather than idiosyncratic bank funding problems. However, notwithstanding these dynamics, the results reported in Table 11 show that the point estimates are relatively stable across periods, even when considering the effect across contract type and education level. In both cases, we find large significant effects on employment levels, mostly concentrated among less educated workers with temporary contracts, which represent almost 4% of total workforce but account for about 30% of total employment effect. The similarity of size of the effects over the two crises is consistent with the evidence showing large real effects of the Lehman and the Greek crises in Italy (for instance, see Bofondi, Carpinelli, and Sette 2017; Bottero et al. 2016; Cingano et al. 2016).32 Table 11 Results by subperiods Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High Time period: 2008:Q1–2011:Q2 All contracts $$\Delta Loan_{t,t-1}$$ 0.380*** 0.183*** 0.165*** 0.0343 (0.113) (0.0627) (0.0579) (0.0209) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.180** 0.0534 0.106** 0.0230* (0.0760) (0.0391) (0.0415) (0.0129) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.196*** 0.126*** 0.0611 0.0108 (0.0713) (0.0458) (0.0379) (0.0163) Observations 1,627,847 1,627,847 1,627,847 1,627,847 1st-stage F-statistic 77.29 77.29 77.29 77.29 Time period: 2009:Q3–2012:Q4 All contracts $$\Delta Loan_{t,t-1}$$ 0.302*** 0.153*** 0.124*** 0.0203 (0.0889) (0.0529) (0.0446) (0.0158) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.167*** 0.0629** 0.0754*** 0.0252*** (0.0513) (0.0276) (0.0275) (0.00926) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.135** 0.0876** 0.0510 –0.00533 (0.0688) (0.0436) (0.0354) (0.0133) Observations 1,955,341 1,955,341 1,955,341 1,955,341 1st-stage F-statistic 140.5 140.5 140.5 140.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High Time period: 2008:Q1–2011:Q2 All contracts $$\Delta Loan_{t,t-1}$$ 0.380*** 0.183*** 0.165*** 0.0343 (0.113) (0.0627) (0.0579) (0.0209) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.180** 0.0534 0.106** 0.0230* (0.0760) (0.0391) (0.0415) (0.0129) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.196*** 0.126*** 0.0611 0.0108 (0.0713) (0.0458) (0.0379) (0.0163) Observations 1,627,847 1,627,847 1,627,847 1,627,847 1st-stage F-statistic 77.29 77.29 77.29 77.29 Time period: 2009:Q3–2012:Q4 All contracts $$\Delta Loan_{t,t-1}$$ 0.302*** 0.153*** 0.124*** 0.0203 (0.0889) (0.0529) (0.0446) (0.0158) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.167*** 0.0629** 0.0754*** 0.0252*** (0.0513) (0.0276) (0.0275) (0.00926) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.135** 0.0876** 0.0510 –0.00533 (0.0688) (0.0436) (0.0354) (0.0133) Observations 1,955,341 1,955,341 1,955,341 1,955,341 1st-stage F-statistic 140.5 140.5 140.5 140.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes The table reports the regression results of the 2SLS estimation of Equation (1) on two different samples that cover the periods 2008:Q1–2011:Q2 (top panel) and 2009:Q3–2012:Q4 (bottom panel). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the alternative credit supply index $$CSI$$, built from a nationwide bank lending regression that include firm$$\times$$quarter fixed effects (see Equation (5) and Section 5.1). The table reports the results for all workers, regardless of their education level (first column), and for those with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education levels based on the ISCED classification. For each of the two subperiods, the table reports separate results for all contracts, open-ended contracts and temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 11 Results by subperiods Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High Time period: 2008:Q1–2011:Q2 All contracts $$\Delta Loan_{t,t-1}$$ 0.380*** 0.183*** 0.165*** 0.0343 (0.113) (0.0627) (0.0579) (0.0209) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.180** 0.0534 0.106** 0.0230* (0.0760) (0.0391) (0.0415) (0.0129) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.196*** 0.126*** 0.0611 0.0108 (0.0713) (0.0458) (0.0379) (0.0163) Observations 1,627,847 1,627,847 1,627,847 1,627,847 1st-stage F-statistic 77.29 77.29 77.29 77.29 Time period: 2009:Q3–2012:Q4 All contracts $$\Delta Loan_{t,t-1}$$ 0.302*** 0.153*** 0.124*** 0.0203 (0.0889) (0.0529) (0.0446) (0.0158) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.167*** 0.0629** 0.0754*** 0.0252*** (0.0513) (0.0276) (0.0275) (0.00926) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.135** 0.0876** 0.0510 –0.00533 (0.0688) (0.0436) (0.0354) (0.0133) Observations 1,955,341 1,955,341 1,955,341 1,955,341 1st-stage F-statistic 140.5 140.5 140.5 140.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High Time period: 2008:Q1–2011:Q2 All contracts $$\Delta Loan_{t,t-1}$$ 0.380*** 0.183*** 0.165*** 0.0343 (0.113) (0.0627) (0.0579) (0.0209) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.180** 0.0534 0.106** 0.0230* (0.0760) (0.0391) (0.0415) (0.0129) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.196*** 0.126*** 0.0611 0.0108 (0.0713) (0.0458) (0.0379) (0.0163) Observations 1,627,847 1,627,847 1,627,847 1,627,847 1st-stage F-statistic 77.29 77.29 77.29 77.29 Time period: 2009:Q3–2012:Q4 All contracts $$\Delta Loan_{t,t-1}$$ 0.302*** 0.153*** 0.124*** 0.0203 (0.0889) (0.0529) (0.0446) (0.0158) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.167*** 0.0629** 0.0754*** 0.0252*** (0.0513) (0.0276) (0.0275) (0.00926) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.135** 0.0876** 0.0510 –0.00533 (0.0688) (0.0436) (0.0354) (0.0133) Observations 1,955,341 1,955,341 1,955,341 1,955,341 1st-stage F-statistic 140.5 140.5 140.5 140.5 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes The table reports the regression results of the 2SLS estimation of Equation (1) on two different samples that cover the periods 2008:Q1–2011:Q2 (top panel) and 2009:Q3–2012:Q4 (bottom panel). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the alternative credit supply index $$CSI$$, built from a nationwide bank lending regression that include firm$$\times$$quarter fixed effects (see Equation (5) and Section 5.1). The table reports the results for all workers, regardless of their education level (first column), and for those with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education levels based on the ISCED classification. For each of the two subperiods, the table reports separate results for all contracts, open-ended contracts and temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. 5.3 Dealing with the large share of zero outcomes Our dependent variable is characterized by a large share of zeros, corresponding to all firm-quarter observations in which the firm does not change its labor force (Table A4). This feature of the data could generate a bias in the estimates and, more important, as the share of zeros varies across subsamples, variations in the extent of the bias could explain part of the heterogeneity of our findings. To better understand the extent to which our results could be driven by differences in the share of zeros across subsamples, we perform two robustness exercises. First, we collapse the data at a yearly frequency. Second, we restrict the sample to firms that have at least one worker with the characteristics (job contracts and demographics) that are the target of the analysis (e.g., the effect on workers with open-ended contracts is studied only on firms with at least one open-ended employee, and the like). Both exercises are aimed at reducing the share of zeros in the dependent variable, but they also imply some drawbacks along other dimensions. In the first case, the decrease in the fraction of zeros is also associated to a significant drop in their absolute number. As the number of firm fixed effects does not change, this entails a significant loss of variability in the data. In the second case, the estimation samples vary across specifications and, therefore, the coefficients are not perfectly comparable across different subsamples. In the first exercise, the share of zeros drops from 79% to 68% in the overall sample, and this trend is even stronger for subcategories of workers. Even though the magnitude of some effects is weaker, and some heterogeneous effects across worker characteristics are not robust, our key findings are qualitatively similar to the baseline ones. Results, reported in Table 12, show that the relative contribution of temporary contracts ($$0.082/0.348 = 0.23$$) is less than half of that estimated with quarterly data ($$0.19/0.36 = 0.52$$), but it is still statistically significant and economically relevant, since it is twice as large as the share of temporary workers in the workforce ($$0.11$$). When interpreting these results it is worth considering that part of the reduced effect on temporary workers could be explained by the fact that moving to a yearly frequency washes out part of the variation in temporary jobs, which is infra-year.33 By contrast, when we look at the heterogeneous effects across education, our baseline results are not confirmed, as low-educated workers account for 38.5% of the total estimated employment adjustment and for 38.7% of the workforce. However, the total employment effect for less educated temporary workers, while much smaller than that estimated with quarterly data, is still more than twice (9%) their share in the workforce. Table 12 Variations at the yearly frequency Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.348*** 0.134*** 0.162** 0.0532** (0.129) (0.0470) (0.0638) (0.0209) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.265** 0.0971** 0.125** 0.0427*** (0.106) (0.0383) (0.0552) (0.0146) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.0816*** 0.0303*** 0.0433*** 0.00942 (0.0280) (0.0113) (0.0153) (0.00780) Observations 715,921 715,921 715,921 715,921 1st-stage F-statistic 578.0 578.0 578.0 578.0 Firm FEs Yes Yes Yes Yes Industry $$\times$$ year FEs Yes Yes Yes Yes Size $$\times$$ year FEs Yes Yes Yes Yes Province $$\times$$ year FEs Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.348*** 0.134*** 0.162** 0.0532** (0.129) (0.0470) (0.0638) (0.0209) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.265** 0.0971** 0.125** 0.0427*** (0.106) (0.0383) (0.0552) (0.0146) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.0816*** 0.0303*** 0.0433*** 0.00942 (0.0280) (0.0113) (0.0153) (0.00780) Observations 715,921 715,921 715,921 715,921 1st-stage F-statistic 578.0 578.0 578.0 578.0 Firm FEs Yes Yes Yes Yes Industry $$\times$$ year FEs Yes Yes Yes Yes Size $$\times$$ year FEs Yes Yes Yes Yes Province $$\times$$ year FEs Yes Yes Yes Yes The table reports the regression results of the 2SLS estimation of Equation (1) on the whole sample, but with data aggregated at the yearly frequency. The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the year $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the year. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in year $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the alternative credit supply index $$CSI$$, built from a nationwide bank lending regression that include firm$$\times$$year fixed effects (see Equation (5) and Section 5.1). The table reports the results for all workers, regardless of their education level (first column), and for those with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education based on the ISCED classification. Results reported in the top panel refer to all contracts, while those in the middle panel refer to open-ended contracts; and the ones in the bottom panel refer to temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 12 Variations at the yearly frequency Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.348*** 0.134*** 0.162** 0.0532** (0.129) (0.0470) (0.0638) (0.0209) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.265** 0.0971** 0.125** 0.0427*** (0.106) (0.0383) (0.0552) (0.0146) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.0816*** 0.0303*** 0.0433*** 0.00942 (0.0280) (0.0113) (0.0153) (0.00780) Observations 715,921 715,921 715,921 715,921 1st-stage F-statistic 578.0 578.0 578.0 578.0 Firm FEs Yes Yes Yes Yes Industry $$\times$$ year FEs Yes Yes Yes Yes Size $$\times$$ year FEs Yes Yes Yes Yes Province $$\times$$ year FEs Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.348*** 0.134*** 0.162** 0.0532** (0.129) (0.0470) (0.0638) (0.0209) Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.265** 0.0971** 0.125** 0.0427*** (0.106) (0.0383) (0.0552) (0.0146) Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.0816*** 0.0303*** 0.0433*** 0.00942 (0.0280) (0.0113) (0.0153) (0.00780) Observations 715,921 715,921 715,921 715,921 1st-stage F-statistic 578.0 578.0 578.0 578.0 Firm FEs Yes Yes Yes Yes Industry $$\times$$ year FEs Yes Yes Yes Yes Size $$\times$$ year FEs Yes Yes Yes Yes Province $$\times$$ year FEs Yes Yes Yes Yes The table reports the regression results of the 2SLS estimation of Equation (1) on the whole sample, but with data aggregated at the yearly frequency. The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the year $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the year. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in year $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the alternative credit supply index $$CSI$$, built from a nationwide bank lending regression that include firm$$\times$$year fixed effects (see Equation (5) and Section 5.1). The table reports the results for all workers, regardless of their education level (first column), and for those with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education based on the ISCED classification. Results reported in the top panel refer to all contracts, while those in the middle panel refer to open-ended contracts; and the ones in the bottom panel refer to temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. The difference across age is relevant, but smaller than in Table 3, while the overall effect on women and foreign workers are not robust to the transition to yearly data. The fact that we do not find results across gender at the yearly frequency could again be due to the fact that women are more likely to be employed with temporary contracts. In fact, when considering the subsample of temporary workers, women are still dis-proportionally affected by the crunch, even if these effects are smaller than those estimated with quarterly data (see annex Table A5). In the second exercise we end up with subsamples with a significant larger number of nonzero observations and, more important, the differences in the relative shares of zero outcomes across subgroups are often mitigated. Our main results are replicated in Tables 13. As expected, coefficients are larger, but we still find a predominant contribution of temporary workers to the overall employment adjustment and, within these job contracts, workers with low and medium education are those affected by the credit crunch. We also confirm the relative larger adjustment borne by women and young workers (see the Online Appendix A-III, Table A6). While this set of results is consistent with a minor role of the bias, we cannot rule out that it may also depend on sample selection, or be a simple mechanical consequence of the fact that we have eliminated from the sample those firms that could not reduce employment in some categories of workers due to the fact that were bound by null stocks. Table 13 Only firms with at least one worker of that type Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.364*** 0.406*** 0.419*** 0.100 (0.111) (0.146) (0.145) (0.0740) Observations 2,459,949 1,212,081 1,230,238 609,171 1st-stage F-statistic 169.5 76.13 48.55 29.60 Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.382*** 0.167** 0.268*** 0.107** (0.121) (0.0764) (0.0932) (0.0507) Observations 1,339,077 1,118,185 1,123,290 571,684 1st-stage F-statistic 67.00 62.45 40.74 23.51 Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.426** 0.301** 0.196* 0.00855 (0.207) (0.152) (0.117) (0.0665) Observations 1,272,214 1,068,597 1,109,633 587,797 1st-stage F-statistic 64.37 61.74 47.48 29.70 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.364*** 0.406*** 0.419*** 0.100 (0.111) (0.146) (0.145) (0.0740) Observations 2,459,949 1,212,081 1,230,238 609,171 1st-stage F-statistic 169.5 76.13 48.55 29.60 Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.382*** 0.167** 0.268*** 0.107** (0.121) (0.0764) (0.0932) (0.0507) Observations 1,339,077 1,118,185 1,123,290 571,684 1st-stage F-statistic 67.00 62.45 40.74 23.51 Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.426** 0.301** 0.196* 0.00855 (0.207) (0.152) (0.117) (0.0665) Observations 1,272,214 1,068,597 1,109,633 587,797 1st-stage F-statistic 64.37 61.74 47.48 29.70 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes The table reports the regression results of the 2SLS estimation of Equation (1) on subsamples of firms with at least one worker of that type (e.g., the effect on workers with open-ended contracts is studied only on firms with at least one open-ended employee, and the like). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the alternative credit supply index $$CSI$$, built from a nationwide bank lending regression that include firm$$\times$$quarter fixed effects (see Equation (5) and Section 5.1). The table reports the results for all workers, regardless of their education level (first column), and for those with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education based on the ISCED classification. Results reported in the top panel refer all contracts, while those in the middle panel to open-ended contracts, and the ones in the bottom panel refer to temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Table 13 Only firms with at least one worker of that type Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.364*** 0.406*** 0.419*** 0.100 (0.111) (0.146) (0.145) (0.0740) Observations 2,459,949 1,212,081 1,230,238 609,171 1st-stage F-statistic 169.5 76.13 48.55 29.60 Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.382*** 0.167** 0.268*** 0.107** (0.121) (0.0764) (0.0932) (0.0507) Observations 1,339,077 1,118,185 1,123,290 571,684 1st-stage F-statistic 67.00 62.45 40.74 23.51 Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.426** 0.301** 0.196* 0.00855 (0.207) (0.152) (0.117) (0.0665) Observations 1,272,214 1,068,597 1,109,633 587,797 1st-stage F-statistic 64.37 61.74 47.48 29.70 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes Dep. var.: $$\Delta Employment_t$$ Education level All Low Medium High All contracts $$\Delta Loan_{t,t-1}$$ 0.364*** 0.406*** 0.419*** 0.100 (0.111) (0.146) (0.145) (0.0740) Observations 2,459,949 1,212,081 1,230,238 609,171 1st-stage F-statistic 169.5 76.13 48.55 29.60 Open-ended contracts $$\Delta Loan_{t,t-1}$$ 0.382*** 0.167** 0.268*** 0.107** (0.121) (0.0764) (0.0932) (0.0507) Observations 1,339,077 1,118,185 1,123,290 571,684 1st-stage F-statistic 67.00 62.45 40.74 23.51 Temporary contracts $$\Delta Loan_{t,t-1}$$ 0.426** 0.301** 0.196* 0.00855 (0.207) (0.152) (0.117) (0.0665) Observations 1,272,214 1,068,597 1,109,633 587,797 1st-stage F-statistic 64.37 61.74 47.48 29.70 Firm FEs Yes Yes Yes Yes Industry $$\times$$ quarter FEs Yes Yes Yes Yes Size $$\times$$ quarter FEs Yes Yes Yes Yes Province $$\times$$ quarter FEs Yes Yes Yes Yes The table reports the regression results of the 2SLS estimation of Equation (1) on subsamples of firms with at least one worker of that type (e.g., the effect on workers with open-ended contracts is studied only on firms with at least one open-ended employee, and the like). The dependent variable $$\Delta Employment_t$$ is defined as the change in employment at the firm level over the quarter $$t$$, for the different types of contracts (open-ended and temporary) and workers (with low, middle, and high education), as labeled in each row and column, respectively. Those flows are divided by the average stock of all firm’s workers over the quarter. Thus, the dependent variable is not a growth rate but a contribution to the aggregate (at the firm level) growth rate. This means that the estimated coefficients cannot be interpreted as elasticities, but they need to be scaled by the relative share of the workers in the workforce (see Table 1). $$\Delta Loan_{t,t-1}$$ is the average change in used loans at the firm level in quarters $$t$$ and $$t-1$$. In the first-stage regressions, the excluded instrument is the alternative credit supply index $$CSI$$, built from a nationwide bank lending regression that include firm$$\times$$quarter fixed effects (see Equation (5) and Section 5.1). The table reports the results for all workers, regardless of their education level (first column), and for those with low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education based on the ISCED classification. Results reported in the top panel refer all contracts, while those in the middle panel to open-ended contracts, and the ones in the bottom panel refer to temporary contracts. Temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. All regressions are based on the full sample and include the same set of time and borrower fixed effects, as listed at the bottom of the table. Fixed effects are constructed based on 30 (2-digit) industries, 7 provinces, and 3 class sizes (firms with fewer than 10 employees, between 10 and 49 employees, and 50 or more employees). Standard errors, clustered at the main bank level, are in parentheses. *** p$$\,{<}\,$$0.01, ** p$$\,{<}\,$$0.05, * p$$\,{<}\,$$0.1. Overall, these exercises show that the large share of zero outcomes and a certain heterogeneity in this dimension may play a role in explaining some of our findings. However, the key result that the employment adjustment following the credit crunch is dis-proportionally concentrated on workers with temporary contracts, as well as other differences across worker characteristics, even if weaker, are confirmed. 6. Conclusions The recent literature on finance and labor has shown that firms reduce employment in response to a credit crunch. Our analysis takes advantage of a novel data set on job contracts and labor market flows for the universe of firms in a large Italian region, to delve into the within-firm personnel dynamics and identify which kind of workers are more likely to be laid off, depending on worker and job contract characteristics. To identify the heterogeneous employment effects of the credit crunch, we rely on loan-level data to build a firm-specific time-varying measure of credit supply restriction, and we control for time-varying demand and productivity shocks using a granular set of borrower fixed effects. Our baseline results confirm that financially constrained firms reduced employment: the point estimate indicates that the average elasticity of employment to a credit supply shock is $$0.36$$. This result is due to an adjustment at the intensive margin, and also to a higher probability of firm closure in response to a reduction in the supply of credit. The adjustment has been strongly differentiated across firms, workers, and job contracts. In particular, the credit crunch has mainly affected women, foreign, and less educated workers with temporary contracts. The effects across workers with temporary contract indicate that labor market regulation is not the only driving force behind the concentration of the employment adjustment on temporary workers. Moreover, these results suggest that firms have adjusted to the credit supply shock in a way that is consistent with a skill upgrading of the labor force. Our findings inform the current debate on the real effects of financial shocks along two main dimensions. First, we show that large credit contractions have distributional effects, as some demographic groups have been more affected than others by the global financial crisis. Second, our analysis indicates that financial shocks could play a cleansing role and foster aggregate productivity gains, given that unskilled workers (and jobs in less productive firms) are more likely to be hit by the credit crunch. In this sense, while credit contractions can have a short-run negative welfare effect, as employment (and investment) falls, in the medium term the labor reallocation toward more educated workers and high-skill occupations could enhance productivity and growth. We wish to thank Isha Agarwal, Ryan Banerjee (discussant), Samuel Bentolila, Tobias Berg (discussant), Mihai Copaciu (discussant), Romaine Duval, Rustom Irani, Andrea Ichino, Francesco Manaresi (discussant), Camelia Minoiu, Holger Mueller (discussant) Michael Neugart, Divya Kirti, Raluca Roman (discussant), Enrico Sette, Nikola Spatafora, and Philip Strahan (the editor); two anonymous referees; and participants to the $$12^{th}$$ NYU Stern/New York Fed Conference on Financial Intermediation (New York, 2017), $$10^{th}$$ Swiss Winter Conference on Financial Intermediation (Lenzerheide, 2017), 2017 AEA Annual Meetings (Chicago, 2017), the Georgetown Center for Economic Research Biennial Conference (Washington DC, 2017), $$7^{th}$$ ICEEE (Messina, 2017), $$2^{nd}$$ IWH-FIN-FIRE workshop on “Challenges on Financial Stability”, Chicago Financial Institutions Conference (Chicago, 2016), $$28^{th}$$ EALE Conference (Ghent, 2016), $$33^{rd}$$ International Symposium on Money, Banking and Finance (Clermont-Ferrand, 2016), the Annual Conference of the Italian Economic Association (Naples, 2015), and the Annual Conferences of the Italian Association of Labour Economists (Cagliari, 2015; Trento, 2016) and at seminars at the Bank of Italy, European University Institute, Federal Reserve Board, International Monetary Fund, University of Oxford and University of Torino for helpful comments and suggestions. We also thank Veneto Lavoro for access to the PLANET data. Financial support from the University of Torino and Compagnia di San Paolo Bank Foundation with project “Skill mismatch: measurement issues and consequences for innovative and inclusive societies” is kindly acknowledged. The views expressed in this paper are those of the authors and do not necessarily represent the views of the International Monetary Fund (IMF), its Executive Board, or IMF management. Supplementary data can be found on The Review of Financial Studies Web site. Footnotes 1 In additional exercises, we show that the effect of the shock is concentrated among firms that entered the crisis with a lower credit rating and higher debt overhang and that have weaker relationships with banks, consistent with the evidence that firm balance sheets play a key role in the propagation of shocks (Giroud and Mueller 2017). We also find that the elasticity of employment to credit supply is especially relevant for micro and small firms, for younger firms, and for those with a lower ex ante labor productivity (see Online Appendix A-III). 2 Using more aggregate data, other papers provide additional evidence of the costs of the financial crisis in terms of reduced employment, in the United States and in Europe (Boeri et al. 2013; Greenstone et al. 2014; Haltenhof et al. 2014; Duygan-Bump et al. 2015). 3 In a related work, Caggese et al. (2016) show that financial constraints distort firms’ firing decisions. Financially constrained firms give more weight to current cash flows than to future ones and therefore decide on whom to fire on the basis of firing costs, rather than considering expected productivity. This hypothesis is confirmed using employer-employee matched data from Sweden that show that financially constrained firms fire relatively more short-tenured workers, who are on average younger, with steeper productivity profiles and lower firing costs, than long-tenured ones. 4 In one of the extensions, we show that results hold even if we build our measure of credit supply on the subsample of firms with multiple bank relationships. Doing so allows us to control for firm-specific time-varying credit demand. 5 In this way, our contribution also relates to and extends the evidence discussed by Caggese and Cuñat (2008), who show that financially constrained firms in Italy have a more volatile labor force and employ a larger proportion of temporary workers than do financially unconstrained firms. 6 One limitation of our data is the lack of information on wages. However, very recent empirical evidence on Europe—and explicitly on Italy—shows that the prevailing labor cost reduction strategy that firms had adopted in response to the Great Recession has worked through the adjustment of quantities rather than prices (Fabiani et al. 2015; Bentolila, Jensen, and Jimenez 2017; Hochfellner et al. 2016), consistently with the presence of downward wage rigidities in regulated labor markets. A further potential constraint of our data is the lack of firm balance-sheet information, a lack that prevents us from controlling for a number of possible drivers of employment decisions. To overcome this limitation, in the empirical analysis we saturate the model with a set of granular fixed effects that capture most of the unobserved time-varying borrower-level heterogeneity. In addition, we match a subsample of relatively larger firms with balance-sheet and income statement data from the CADS database—a proprietary firm-level database owned by Cerved Group S.p.a.—to explore additional sources of firm-level heterogeneity, and assess the effect of the credit crunch on capital accumulation (see Online Appendix A-III). 7 We do not (explicitly) include interest rates when examining the impact of credit conditions on firm employment for two main reasons. First, data on interest rates are collected only for a subsample of banks that exclude the majority of small and local banks and this would have entailed a severe reduction of observations and the dismissal of our census analysis perspective. Second, one may reasonably argue that bank policies on prices are correlated with those on quantities and that utilized loans—which we use in our analysis—reflect both granted loans and (unobserved) price effects. 8 To construct our measure of credit supply, we use data drawn from the Bank of Italy Supervisory Report (SR) database. Specifically, we use confidential data on outstanding loans extended by Italian banks to the firms in the local credit markets (i.e., provinces) to estimate time-varying bank lending policies. 9 We also remove temp agencies, caregivers and house cleaners from our sample. The reason for temp agencies is that we cannot distinguish between the internal staff and the workers leased to other firms, and since temp agency workers are also included within the employed workforce of the firms they are leased to, retaining temp agencies would results in a duplication of flow records. Caregivers and house cleaners, instead, are excluded because in most cases they appear as self-employed if not individual firms. In the latter case, they would mistakenly increase the number of actual firms. Moreover, when registered as employees, they are typically employed by households, rather than by firms. 10 We measure loan growth using utilized loans rather than granted loans because the former captures rationing in terms of both a reduction in granted loans (i.e., quantity side) and/or an increase in interest rates (i.e., price effects). 11 The dependent variable shows a high variability, with a large number of negative and positive changes, but it has a large share of zeros (corresponding to all firm-quarter observations in which the firm does not change its labor force), which could generate a bias in the estimate. In some circumstances, variations in the share of zeros across subsamples, could explain part of the heterogeneity of our findings. We extensively discuss this issue in Section 5.3. 12 In Section 5 we will show that our key results hold if we exclusively consider the contemporaneous change in loans, or the average change over three quarters. 13 On the one hand, low performing firms can be more likely to demand/receive less credit and to contract the labor force, inducing an upward bias in the OLS estimates. On the other hand, the OLS could be downward biased because of “evergreening” practices, so that firms under stress would reduce their employment, but at the same time receive additional credit from their banks (Peek and Rosengren 2005). 14 Provinces correspond to NUTS 3 Eurostat classification (a geography entity similar to U.S. counties), and, according to the supervisory authority, they represent the “relevant” market in banking (see also Guiso et al. 2004). 15 Moreover, data show that the large drop in credit supply conditions from the beginning of the financial crisis onward was mostly concentrated among large banks, consistent with the fact that those banks were more exposed to the liquidity drought in interbank markets. 16 The “diffusion indexes” reflects subjective assessments of the lender on the relative importance of demand and supply factors in explaining the lending patterns. Technically, the diffusion index is the (weighted) difference between the share of banks reporting that credit standards have been tightened and the share of banks reporting that they have been eased. 17 While, by definition, the set of observables cannot include all possible firm characteristics, we argue that it is difficult to identify firm characteristics which are correlated with the credit supply index, while being orthogonal to the variables listed in Tables 2 and A12. Also, one could argue that the credit supply index is spatially autocorrelated—for instance, because some banks control large market shares in certain areas. This is indeed the case in our data, as shown by Figure A1 in the Online Appendix and more formally by the Moran index calculated on $$CSI$$. However, once we control for the standard set of fixed effects, the Moran index does not show evidence of spatial autocorrelation in $$CSI$$ and the residuals of our baseline regression are also not spatially autocorrelated. 18 An alternative identification strategy is the one proposed by Amiti and Weinstein (2017), who identify the bank shocks (i.e., time-varying bank fixed effects) through a regression on the dynamic of loans at the firm level, exploiting information from the subsample of firms who borrow from multiple banks. However, we believe that their approach is less suitable for our case since only about one-third of firms in our sample borrow from more than one bank at the beginning of the sample period. However, in Section 5.1 we discuss results obtained identifying the bank fixed effects $$\delta_{bt}$$ in a regression at the firm level with time-varying firm fixed effect, on a subsample of borrowers with multiple bank relationships. 19 The exclusion of Veneto provinces from the estimation of bank lending policies leads to the exclusion of only one bank (accounting for less than 0.1% of loans granted to all firms residing in Veneto), for which we were not able to estimate the national lending policy. Therefore, this strategy does not affect the representativeness of our sample, but it strongly reinforces the exogeneity of the instrument. Veneto represents about 8% of total loans granted by the median bank active in the region. 20 More precisely, the bank corresponds to the firm’s main bank. Results are robust to alternative levels of clustering, which provide more conservative estimates compared to simple heteroscedasticity robust standard errors. In particular, we estimate bank lending policies with data at the bank-province-sector level to construct the credit supply index, and in annex Table A2 we show that clustering at that level leads to very similar standard errors. 21 This elasticity is roughly twice as large as the one estimated by Cingano et al. (2016) for a sample of larger Italian firms, but over a different time span, confirming that focusing on the universe of firms helps provide a more precise estimate of the employment effect of a credit restriction. 22 In other words, the dependent variable is calculated as the ratio between the job flows for a given category of contracts or workers—which we retrieve from PLANET—and the average stock of total workers ($$0.5 \times X_{it_1} + 0.5 \times X_{it_0}$$, as defined at the denominator of Equation (2)). 23 Since firms do not have to pay dismissal costs upon termination of temporary contracts, they typically employ temporary workers as a buffer stock, to deal with expected or unexpected fluctuations in demand or in financial conditions. Indeed, recourse to temporary contracts is known to be more cyclical than the use of open-ended contracts (García Serrano 1998; Goux et al. 2001). 24 We also perform a test to confirm that the relative contribution of temporary and open-ended contracts to the estimated change in employment is statistically different from their share in the workforce, as listed in Table 1. The rationale for this test is that, if employment losses were random across contract type, each contract contribution to overall loss should mirror its share in the total workforce. The same reasoning also applies for other subgroups of workers; see the discussion that follows. 25 In the Online Appendix A-III, we discuss an additional set of results that exploit cross-firm heterogeneity. We find that smaller, younger, and less productive firms and those with higher debt overhang and weaker bank-firm relationships have been more vulnerable to the (negative) impact of the credit crunch on employment. 26 The difference between the relative contribution of low-educated workers to the estimated change in employment ($$0.51$$) is statistically different from their share in the labor force ($$0.39$$). As education may not perfectly overlap with the skill content of jobs, we replicate the analysis by skill level by directly looking at the skill content of each occupation. The findings based on this different measure of skill level are stronger than those based on the education level: the effect is predominantly concentrated on low-skill occupations, which represent about 15% of jobs, but account for about 43% of the total effect of the credit contraction. Results are not shown but are available on request. 27 While this difference is statistically significant, we cannot exclude that some of the penalty for foreign workers comes from sheer discrimination. For instance, it has been documented that economic downturns favor racial prejudice and lead to worse labor market outcomes for minorities (Johnston and Lordan 2016). 28 When considering foreign workers, their relative contribution to the change in employment is significantly larger than their share in the labor force for both types of contracts, even though the coefficient for temporary contracts is imprecisely estimated. 29 To further rule out the hypothesis that EPL is driving our findings, in robustness exercises we exploit the fact that in Italy, during the observed period, the strength of EPL differs in a quite substantive way between firms across a threshold of 15 employees, with smaller firms facing weaker EPL and having more flexibility in adjusting their labor force. We replicate our key analysis on a subsample of firms with fewer than 15 employees, and we still find that out results hold (see Online Appendix A-III, Table A15). 30 The average firm size is nearly 3,000 in Chodorow-Reich (2014) and about 25 in Bentolila, Jensen, and Jimenez (2017), whereas it is around 6 in our case, because we are able to observe the universe of firms. 31 The prevalence of micro-enterprises could also undermine the generality of our findings to other setting where the presence of micro-firms is lower. To deal with this issue, we run the analysis on a subsample comprising firms with at least 3 employees (see annex Table A3), and we find results in line with those on the whole sample, suggesting that micro-firms are not those driving our findings. 32 We also find that the differential effects across gender, nationality, and age hold in the two subsamples. Results are not shown, but they are available on request. 33 In 2012, 46% of temporary contracts had a duration shorter than 1 month, 19.6% has a duration of 2-3 months, 30.9% had duration of 4-12 months, and only 3.6% of temporary contracts expired after more than one year. References Abuka, C., Alinda, R. K. Minoiu, C. Peydro, J.-L. and Presbitero. A. F. 2017 . Financial development and monetary policy: Loan applications, rates, and real effects. Discussion Paper , Center for Economic and Policy Research (CEPR) , London . Acharya, V. V., Eisert, T. Eufinger, C. and Hirsch. C. 2016 . Real effects of the sovereign debt crisis in Europe: evidence from syndicated loans. Mimeo, NYU . Adelino, M., Schoar, A. and Severino. F. 2015 . House prices, collateral, and self-employment. Journal of Financial Economics 117 : 288 – 306 . Google Scholar CrossRef Search ADS Altman, E. I. 1968 . Financial ratios, discriminant analysis and the prediction of corporate bankruptcy. Journal of Finance 23 : 589 – 609 . Google Scholar CrossRef Search ADS Amiti, M., and Weinstein. D. E. 2011 . Exports and financial shocks. Quarterly Journal of Economics 126 : 1841 – 77 . Google Scholar CrossRef Search ADS Amiti, M., and Weinstein. D. E. Forthcoming. How much do idiosyncratic bank shocks affect investment? Evidence from matched bank-firm data. Journal of Political Economy . Auer, R., and Ongena. S. 2016 . The countercyclical capital buffer and the composition of bank lending. Unpublished Manuscript . Barbosa, L., Bilan, A. and Celerier. C. 2017 . Credit supply shocks and human capital: Evidence from a change in accounting norms. Working Paper . Beck, T., Demirgüç-Kunt, A. and Maksimovic. V. 2008 . Financing patterns around the world: Are small firms different? Journal of Financial Economics 89 : 467 – 87 . Google Scholar CrossRef Search ADS Benmelech, E., Bergman, N. K. and Seru. A. 2015 . Financing labor. Working Paper . Bentolila, S., Jensen, M. and Jimenez. G. When credit dries up: Job losses in the great recession. Journal of the European Economic Association. Advance Access published August 17 , 2017 , 10.1093/jeea/jvx021 . Berg, T. 2016 . Got rejected? Real effects of not getting a loan. Working Paper , European Central Bank . Google Scholar CrossRef Search ADS Boeri, T., Garibaldi, P. and Moen. E. R. 2013 . Financial shocks and labor: Facts and theories. IMF Economic Review 61 : 631 – 63 . Google Scholar CrossRef Search ADS Bofondi, M., Carpinelli, L. and Sette. E. 2017 . Credit supply during a sovereign debt crisis. Journal of the European Economic Association. Advance Access published August 17 , 2017 , 10.1093/jeea/jvx020 . Bottero, M., Lenzu, S. and Mezzanotti. F. 2016 . Sovereign debt exposure and the bank lending channel: impact on credit supply and the real economy. Working Paper , Bank of Italy . Buera, F., Fattal-Jaef, R. and Shin. Y. 2015 . Anatomy of a credit crunch: From capital to labor markets. Review of Economic Dynamics 18 : 101 – 17 . Google Scholar CrossRef Search ADS Burlon, L., Fantino, D. Nobili, A. and Sene. G. 2016 . The quantity of corporate credit rationing with matched bank-firm data. Working Papers , Bank of Italy . Google Scholar CrossRef Search ADS Caballero, R. J., and Hammour. M. L. 1994 . The cleansing effect of recessions. American Economic Review 84 : 1350 – 68 . Caggese, A., and Cuñat. V. 2008 . Financing constraints and fixed-term employment contracts. Economic Journal 118 : 2013 – 2046 . Google Scholar CrossRef Search ADS Caggese, A., Cuñat, V. and Metzger. D. 2016 . Firing the wrong workers: Financing constraints and labor misallocation. Mimeo, London School of Economics . Chodorow-Reich, G. 2014 . The employment effects of credit market disruptions: Firm-level evidence from the 2008–9 financial crisis. Quarterly Journal of Economics 129 : 1 – 59 . Google Scholar CrossRef Search ADS Cingano, F., Manaresi, F. and Sette. E. 2016 . Does credit crunch investments down? New evidence on the real effects of the bank-lending channel. Review of Financial Studies 29 : 2737 – 73 . Google Scholar CrossRef Search ADS Degryse, H., De Jonghe, O. Jakovljević, S. Mulier, K. and Schepens. G. 2016 . The impact of bank shocks on firm-level outcomes and bank risk-taking. Working Paper . Duygan-Bump, B., Levkov, A. and Montoriol-Garriga. J. 2015 . Financing constraints and unemployment: Evidence from the Great Recession. Journal of Monetary Economics 75 : 89 – 105 . Google Scholar CrossRef Search ADS Ersahin, N., and Irani. R. 2016 . Collateral values and corporate employment. Mimeo, University of Illinois at Urbana-Champaign . Google Scholar CrossRef Search ADS Fabiani, S., Lamo, A. Messina, J. and Room. T. 2015 . European firm adjustment during times of economic crisis. IZA Journal of Labor Policy 4 : 4 – 24 . Google Scholar CrossRef Search ADS Forsythe, E. 2016 . Why don’t firms hire young workers during recessions? Working Paper , University of Illinois . Foster, L., Grim, C. and Haltiwanger. J. 2016 . Reallocation in the great recession: Cleansing or not? Journal of Labor Economics 34 : S293 – S331 . Google Scholar CrossRef Search ADS Gambacorta, L., and Mistrulli. P. E. 2014 . Bank heterogeneity and interest rate setting: What lessons have we learned since Lehman Brothers? Journal of Money, Credit and Banking 46 : 753 – 78 . Google Scholar CrossRef Search ADS García Serrano, C. 1998 . Worker turnover and job reallocation: the role of fixed-term contracts. Oxford Economic Papers 50 : 709 – 25 . Google Scholar CrossRef Search ADS Gertler, M., and Gilchrist. S. 1994 . Monetary policy, business cycles, and the behavior of small manufacturing firms. Quarterly Journal of Economics 109 : 309 – 40 . Google Scholar CrossRef Search ADS Giroud, X., and Mueller. H. M. 2017 Firm leverage, consumer demand, and unemployment during the great recession. Quarterly Journal of Economics 132 : 271 – 316 . Gobbi, G., and Sette. E. 2014 . Do firms benefit from concentrating their borrowing? Evidence from the great recession. Review of Finance 18 : 527 – 60 . Google Scholar CrossRef Search ADS Goux, D., Maurin, E. and Pauchet. M. 2001 . Fixed-term contracts and the dynamics of labour demand. European Economic Review 45 : 533 – 52 . Google Scholar CrossRef Search ADS Greenstone, M., Mas, A. and Nguyen. H.-L. 2014 . Do credit market shocks affect the real economy? Quasi-experimental evidence from the great recession and ‘normal’ economic times. Working Paper , National Bureau of Economic Research . Guiso, L., Sapienza, P. and Zingales. L. 2004 . Does local financial development matter? Quarterly Journal of Economics 119 : 929 – 69 . Google Scholar CrossRef Search ADS Haltenhof, S., Lee, S. J. and Stebunovs. V. 2014 . The credit crunch and fall in employment during the great recession. Journal of Economic Dynamics and Control 43 : 31 – 57 . Google Scholar CrossRef Search ADS Haltiwanger, J., Jarmin, R. S. and Miranda. J. 2013 . Who creates jobs? Small versus large versus young. Review of Economics and Statistics 95 : 347 – 61 . Google Scholar CrossRef Search ADS Hershbein, B., and Kahn. L. B. 2016 . Do recessions accelerate routine-biased technological change? Evidence from vacancy postings. Working Paper , Upjohn Institute . Hochfellner, D., Montes, J. Schmalz, M. and Sosyura. D. 2016 . Winners and losers of financial crises: Evidence from individuals and firms. Working Paper , University of Michigan . Hoynes, H., Miller, D. L. and Schaller. J. 2012 . Who suffers during recessions? Journal of Economic Perspectives 26 : 27 – 48 . Google Scholar CrossRef Search ADS Jiménez, G., Mian, A. R. Peydro, J.-L. and Saurina. J. 2014 . The real effects of the bank lending channel. Working Paper . Johnston, D. W., and Lordan. G. 2016 . Racial prejudice and labour market penalties during economic downturns. European Economic Review 84 : 57 – 75 . Google Scholar CrossRef Search ADS Kalemli-Ozcan, S., Laeven, L. and Moreno. D. 2015 . Debt overhang on Europe: Evidence from firm-bank-sovereign linkages. Working Paper , University of Maryland . Khwaja, A. I., and Mian. A. 2008 . Tracing the impact of bank liquidity shocks: Evidence from an emerging market. American Economic Review 98 : 1413 – 42 . Google Scholar CrossRef Search ADS Mian, A., and Sufi. A. 2014 . What explains the 2007–2009 drop in employment? Econometrica 82 : 2197 – 223 . Google Scholar CrossRef Search ADS Moscarini, G., and Postel-Vinay. F. 2012 . The contribution of large and small employers to job creation in times of high and low unemployment. American Economic Review 102 : 2509 – 39 . Google Scholar CrossRef Search ADS Pagano, M., and Pica. G. 2012 . Finance and employment. Economic Policy 27 : 5 – 55 . Google Scholar CrossRef Search ADS Paravisini, D., Rappoport, V. Schnabl, P. and Wolfenzon. D. 2015 . Dissecting the effect of credit supply on trade: Evidence from matched credit-export data. Review of Economic Studies 82 : 333 – 59 . Google Scholar CrossRef Search ADS Peek, J., and Rosengren. E. S. 2005 . Unnatural selection: perverse incentives and the misallocation of credit in Japan. American Economic Review 95 : 1144 – 66 . Google Scholar CrossRef Search ADS Petrosky-Nadeau, N. 2013 . TFP during a credit crunch. Journal of Economic Theory 148 : 1150 – 78 . Google Scholar CrossRef Search ADS Popov, A., and Rocholl. J. 2016 . Do credit shocks affect labor demand? Evidence from employment and wages during the financial crisis. Journal of Financial Intermediation. Advance Access published November 7 , 2016 , 10.1016/j.jfi.2016.10.002 . Presbitero, A. F., Udell, G. F. and Zazzaro. A. 2014 . The home bias and the credit crunch: A regional perspective. Journal of Money, Credit, and Banking 46 : 53 – 85 . Google Scholar CrossRef Search ADS Reder, M. W. 1955 . The theory of occupational wage differentials. American Economic Review 45 : 833 – 52 . Schivardi, F., Sette, E. and Tabellini. G. 2017 . Credit misallocation during the European financial crisis. CEPR Discussion Paper , Center for Economic and Policy Research . Siemer M. 2016 . Employment effects of financial constraints during the Great Recession. Board of Governors of the Federal Reserve System , June . Google Scholar CrossRef Search ADS Staiger D., and Stock. J. H. 1997 . Instrumental variables regression with weak instruments. Econometrica 65 : 557 – 86 . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of The Society for Financial Studies. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Review of Financial Studies Oxford University Press

# Banks, Firms, and Jobs

, Volume Advance Article (6) – Jan 16, 2018
44 pages

/lp/ou_press/banks-firms-and-jobs-OE2lSV02I0
Publisher
Oxford University Press
Abstract We analyze the heterogeneous employment effects of financial shocks using a rich data set of job contracts, matched with the universe of firms and their lending banks in one Italian region. To isolate the effect of the financial shock, we construct a firm-specific time-varying measure of credit supply. The preferred estimate indicates that the average elasticity of employment to a credit supply shock is $$0.36$$. Adjustment affects both the extensive and the intensive margins and is concentrated among workers with temporary contracts. We also examine the heterogeneous effects of the credit crunch by education, age, gender and nationality. Received January 27, 2017; editorial decision December 1, 2017 by Editor Philip Strahan. Authors have furnished an Internet Appendix, which is available on the Oxford University Press Web site next to the link to the final published paper online. Introduction In the aftermath of the global financial crisis, a severe credit crunch has had long-lasting consequences on a number of advanced economies, where unemployment rates have increased markedly. The labor market effects of the crisis have not been uniform, and, in particular, young and less educated workers have been particularly hit by the crisis (Hoynes et al. 2012). Although these outcomes raise important distributional concerns, it has been argued that crises could also have a “cleansing” effect to the extent that the least productive jobs and firms are the ones relatively more affected by the financial shock (Caballero and Hammour 1994; Petrosky-Nadeau 2013). The effects of the global financial crisis on the labor market have triggered a renewed interest in the relationship between finance and employment (Pagano and Pica 2012) and, specifically, a renewed interest in the effect that credit supply shocks have on firms’ employment decisions (Chodorow-Reich 2014; Buera et al. 2015; Duygan-Bump et al. 2015). While this literature provides original insights into the effects that financial crises have on aggregate employment at the firm or state level, it is generally silent about within-firm dynamics and labor reallocation. For instance, little is known about the impact of a decline in firm financing on different types of jobs, even though a differential impact of the crisis across demographic groups would have distributional implications. Moreover, the employment adjustment within firms—between workers and jobs characterized by a different skill set—can have an effect on aggregate productivity. We contribute to this strand of literature by zooming in on the employment dynamics within the firm and by providing a series of novel findings on how firms adjust the level and composition of the labor force in response to credit shocks. In particular, we focus on worker education to test whether the contraction in employment during the global financial crisis has been associated with a skill upgrading of the workforce, at the firm level. Knowing which jobs and workers are more exposed to the real effects of large financial shocks provides useful insight into better understanding how firms reorganize themselves at times of crisis and can inform the debate on the distributional consequences and the possible cleansing effect of financial crises. We run our analysis thanks to the availability of an original and extremely rich data set, that draws on an administrative archive that collects daily information on individual job contracts and labor market flows. The data set covers the universe of firms, including micro-enterprises, in an Italian region, matched with their lending banks through the Italian Credit Register. This is an important feature of our data given that bank credit is very often the only source of external financing for micro and small enterprises. We end up with a quarterly data set of about 200,000 firms, spanning the period from 2008:Q1 to 2012:Q4, for which, thanks to the degree of granularity of the data, we can go beyond the standard job destruction/job creation dichotomy to investigate differential responses to a credit supply shocks across firms, workers, and job contracts. We find that a 10% supply-driven credit contraction reduces employment by 3.6%. This effect is the result of adjustments at the intensive and extensive margins, is concentrated among workers with temporary contracts, and mostly occurred through increased outflows rather than decreased inflows. These results are in line with the existence of a “dual” labor market in which temporary contracts absorb a large part of the employment volatility. The reduction in employment is concentrated among relatively less educated individuals and mostly happened by allowing temporary contracts to expire. By contrast, less educated workers with open-ended contracts are almost unaffected by tighter firms’ financing constraints, possibly because of higher firing costs and a rigid employment protection legislation (EPL). Even though skill upgrading strategies are heavily shaped by contract regulation, our results are not exclusively driven by high EPL: when focusing on a subsample of small firms for whom firing costs are lower, we still find that the adjustment is primarily borne by less educated workers with temporary contracts. These differential effects are mainly driven by the adjustment at the intensive margin, whereas the effects on employment due to firm exit are more homogeneous across contracts and workers. We also find evidence suggesting that women and foreign workers are hit disproportionately more by the credit shock, regardless of the kind of job contract, while the stronger effect on young workers reflects their higher propensity to be hired with temporary contracts.1 From a more general perspective, it could still be the case that workers who are more likely to lose their job in our sample could be hired by nonbanked firms. However, when we aggregate employment and loan outcomes for all firms in the region at the province-industry-quarter level, we still find a negative impact of the credit crunch (both from an economic and statistical point of view) on employment, suggesting that the contraction in employment has not been offset by firms without banking relationships. This paper contributes to the growing literature on the real effect of credit supply shocks (Amiti and Weinstein 2011, 2017; Cingano et al. 2016; Paravisini et al. 2015) and is closely related to the recent contributions that investigate the effects of financial shocks on employment outcomes at the firm level (Barbosa et al. 2017; Benmelech et al. 2015; Bentolila, Jensen, and Jimenez 2017; Berg 2016; Caggese et al. 2016; Chodorow-Reich 2014; Ersahin and Irani 2016; Giroud and Mueller 2017; Hochfellner et al. 2016; Popov and Rocholl 2016; Siemer 2016).2 Drawing on micro-level data sets, these studies consistently show that a tightening of the credit supply leads to a contraction of the workforce. The analysis by Bentolila, Jensen, and Jimenez (2017) has the unique feature of being based on loan-level data from a credit register. Relying on the differences in bank health at the beginning of the financial crisis, the paper shows that firms exposed to weak banks contracted employment by 2.8 percentage points more than firms that were borrowing from healthier lenders, and results are able to explain about a fourth of the fall in aggregate employment in Spain between 2007 and 2010. Also, their analysis uncovers that job losses have been mostly borne by temporary employees, while wages adjusted only marginally. Hochfellner et al. (2016) use employer-employee matched data for a sample of German firms to look at how individual characteristics affect labor outcomes. The identification strategy hinges on differences in firm location, distinguishing between firms that are located in one of the seven federal states in which the major bank was one of the five Landesbanks with significant exposure to the U.S. mortgage crisis and firms that are located elsewhere. In addition to confirming the aggregate negative effect of credit contraction on employment, Hochfellner et al. (2016) show that workers in firms which have been exposed to a negative credit shock experience significant earning losses and an increase in the unemployment spell. They also find that unskilled, less educated and less experienced workers are the most affected by the credit shock.3 While both these studies limit their analysis to medium-sized and large firms, Siemer (2016) uses confidential firm-level employment data from the U.S. Bureau of Labor Statistics for the universe of U.S. firms, but relies on industry-level differences in external financial dependence to identify the effects of financial constraints on employment and firm dynamics. His results show that financing constraints reduce employment growth in small firms by 5 to 10 percentage points relative to large firms, but they are silent on within-firm heterogeneity. Our analysis has the advantage of bringing together three key elements which in previous studies have been considered separately. First, the availability of loan-level data (instead of aggregate credit data) allows us to identify the bank lending channel at the firm-level. Moreover, those data make it possible to control for credit demand and productivity shocks at a granular level, with a set of firm, time, and firm cluster$$\times$$time fixed effects, which absorb firm-specific time invariant demand shifters and time-varying demand shocks that are common to a narrowly defined cluster of borrowers. The matched bank-firm data also allow us to extend the identification strategy of Greenstone et al. (2014) and construct an exogenous firm-specific time-varying measure of bank credit supply, which gives us more precise estimates than the ones obtained with more aggregate data. We start by estimating time-varying nationwide bank lending policies that are purged of local loan demand (and of any other province-sector-quarter level idiosyncratic shocks). Then we build a credit supply variable at the firm level using banks’ loan share to a given firm as weights. We discuss different arguments to motivate the exogeneity of our instrument and we show that it is strongly correlated with loan growth at the firm level.4 Second, thanks to contract-firm-bank-matched data, we can investigate heterogeneous responses to a financial shock across workers, job contracts, and firms. In particular, we can exploit differences across contract types and look at the intersection between demographic characteristics (education, age, gender and nationality) and job contracts, to assess which dimensions matter more for firm’s employment decisions.5 Third and finally, our analysis covers the universe of firms. While there is a wide consensus on the fact that smaller firms rely more on bank financing, the existing evidence rarely focuses on a representative sample of small firms. Our data, on the contrary, include the universe of individual and micro enterprises, and this allows us to have a more precise (and larger) estimate of the employment effect of financial shocks. 1. Data 1.1 Veneto as a representative case study Our analysis relies on unparalleled loan-level information about the entire population of workers, firms and financial intermediaries operating in Veneto, a large Italian region with a population of 4.9 million individuals and a workforce of 2.2 million workers. According to the National Institute of Statistics data, the region accounts for roughly 9% of the Italian value added and of total employment. A key feature for our analysis is that Veneto can be considered as a self-contained labor market. About 97% of the workers who reside in the region have their workplace in a municipality within the region, and migration to other regions is a negligible phenomenon at the aggregate level (0.4% of the population per year); moreover, both figures are substantially stable in the temporal window considered in the analysis. As a result, it is unlikely that our results will be biased by dismissed workers finding jobs out of region. Veneto shares with Italy a large prevalence of small firms (Figure 1, panel a): 94% of firms in the region have fewer than 10 employees (57% have at most one employee). The productive structure is also fairly similar to the national one (Figure 1, panel b), and the service and industrial sectors accounts for 56% and 43% of total employment, respectively, with the share of the industrial sector being slightly larger than in the rest of Italy. Figure 1 View largeDownload slide External validity: Firm distribution across size and sectors in Veneto and Italy The figure further elaborates ISTAT data (census 2011). Figure 1 View largeDownload slide External validity: Firm distribution across size and sectors in Veneto and Italy The figure further elaborates ISTAT data (census 2011). In terms of the banking system, in 2012 in Veneto there were about 120 banks, with small local banks accounting for nearly 20% of business loans. The degree of financial development, as measured by the number of branches per inhabitants, is higher with respect to the national average (Figure 2, panel a). Aggregate lending to nonfinancial corporations followed a similar dynamic in Veneto and Italy (Figure 2, panel b). Figure 2 View largeDownload slide External validity: Bank penetration and lending in Veneto and Italy The figure further elaborates data from Bank of Italy. Figure 2 View largeDownload slide External validity: Bank penetration and lending in Veneto and Italy The figure further elaborates data from Bank of Italy. Veneto is hence very well representative of the Italian situation, which in turn represents an extremely interesting case studies for at least two reasons: first, Italian firms mostly rely on bank credit for their business activities, and more than other firms in the Euro area (Figure 3, panel a); second, small firms (fewer than 10 employees) are the most indebted, and the Italian productive structure is strongly biased toward small production units (Figure 3, panel b). Figure 3 View largeDownload slide Bank financing in Italy across firm size The figure further elaborates data from the Survey on the Access to Finance of Enterprises (SAFE, European Central Bank), Bank of Italy, PLANET, and ASIA. Debt per employee is measured in thousands of euro. Figure 3 View largeDownload slide Bank financing in Italy across firm size The figure further elaborates data from the Survey on the Access to Finance of Enterprises (SAFE, European Central Bank), Bank of Italy, PLANET, and ASIA. Debt per employee is measured in thousands of euro. 1.2 The contract-firm-bank-matched data Our data set brings together an extremely rich set of information coming from different administrative sources. In the following we provide an overview of the construction and structure of data set, while more detailed information are discussed in the annex A-I. Daily labor market flows from the regional public employment service are indeed matched to stock information from the national social security administration and to the Italian credit register maintained by the Bank of Italy using firm-level unique identifiers, namely their VAT numbers. These feature of the data guarantees at the same time wide population coverage, high information reliability and a nearly total frequency of success in the matching procedure. The bulk of labor market information comes from PLANET, an administrative data set of daily labor market flows maintained by the regional employment agency Veneto Lavoro. PLANET builds on the obligation for firms operating in Italy to notice the national and local employment agencies about all labor market transitions for which they are held responsible, including hires, firings and transformations of individual employment arrangements (e.g., from full-time to part-time, from temporary to permanent, and the like). Firm-level observables include geographical location and sector (5-digit NACE code), while worker information covers gender, age, nationality, occupation (5-digit ISCO code), type of contract (44 different employment arrangement), educational attainment (13 categories), time schedule (full-time or vertical, horizontal or mixed part-time), and reasons for separation from the firm. To overcome limitations in terms of labor market stocks, PLANET is complemented with information from ASIA, the archive of active firms maintained by the National Statistical Institute (ISTAT) with register data from the Social Security Administration. ASIA provides yearly data about firms whose economic activity spans for at least six months within a calendar year. To our purposes, ASIA adds information on firm size and on characteristics of those firms who are not interested by any job flows or transitions in our sample period. More specifically, we consider the stock in the first year in which we observe the firm, and we reconstruct the stock forward using information on workers inflows and outflows. The purpose of this exercise is to guarantee consistency between flows and stocks and, more importantly, to have quarterly stock data.6 To obtain a firm-specific measure of credit availability, we use information from the Credit Register (CR) database, managed by the Bank of Italy, on the credit extended to each firm in each quarter. For each borrower, banks have to report to the Register, on a monthly basis, the amount of each loan—granted and used—for all loans exceeding a minimum threshold (75,000 euro until December 2008, 30,000 euro afterward), plus all nonperforming loans. Given the low threshold, these data can be taken as a census.7 Data also contain a breakdown by type of the loan (e.g., credit lines, credit receivables and fixed-term loans). From CR we essentially draw two kind of information. First, borrower’s outstanding loans (from all banks operating in Italy) at the end of each quarter: we consider the total amount instead of the different types of loans because banks and borrowers may endogenously change the composition of loans in reaction to shocks to the credit market. Second, the bank market share for each borrower at the beginning of the period, that we use to construct the instrumental variable (see Section 2.2).8 1.3 Sample selection and the final data set All data sources are merged together using VAT numbers as univocal firm identifiers. Genuine nonmatches between PLANET and ASIA are possible, and are due to two reasons: very short-lived firms (less than a semester in a calendar year) are not recorded in ASIA, while firms with a very stable employed workforce (meaning no changes in both the intensive and the extensive margins, including the type of contract) do not appear in PLANET. None of the two entails any limitation to our purposes, as (1) the stock of employed workforce for very short-lived firms can be easily induced from workers’ flows, and (2) the worker flows in stable firms are by definition null. Moreover, all firms with loan information are also present in ASIA, so extremely short-lived firms fall beyond the scope of the analysis. Thus, we include all firms that are not in PLANET but are in the firm register, and we assume that inflows and outflows for those firms are zero. This grants that truly unsuccessful matches are infrequent and largely due to misreporting of VAT numbers by either the firms or the statistical offices maintaining the single sources, an occurrence that we can safely assume to be random and—due to the extremely large sample size—almost irrelevant from a statistical standpoint. The selection of the sample is driven by two main reasons. First, although the available time series cover a longer period, we narrow our focus on the years from 2008 to 2012 (the last available year in most sources at the time of our analysis). The reason is that until 2007 the obligation for firms to notice hires and firings (from which PLANET originates) concerned dependent workers only and occurred largely through paper documents. The first limitation resulted in an incomplete coverage of labor market flows, insofar as independent contractors and disguised self-employees—widely spread in the Italian labor market and at high risk to represent a buffer stock of employment during downturns—were not observed in the data. The second limitation entailed a nonnegligible delay of data completion. Both were overcome during 2007, when digital notice became compulsory for all workers, including independent ones. Second, we focus on the private nonfinancial nonprimary sectors. The reasons are self-evident. Employment in the public sector depends on different rationales that include macroeconomic stabilization, budget control and the supply of public services, and its funding relies to a great deal on out-of-market sources (taxes). The agriculture sector in turn is highly subsidized all over the EU and a credit crunch from the private sector may be overcome by financial resources that we cannot observe at the micro level. Finally, credit flows within the financial sector often respond to different factors than flows from banks to nonfinancial corporations.9 After a process of data cleansing, the final sample includes nearly 440,000 firms of which about 200,000 have bank relationships. 1.4 Descriptive statistics The firms included in the sample are predominantly micro and small enterprises, reflecting the structure of the Italian industry. This distribution is consistent with Census data both in terms of firms and employees (Figure 4). Over the sample period 2008–2012, the number of employees declines by nearly 90,000 units, and the number of firms records a significant drop too. These trends mimic the aggregate data from the National Institute of Statistics (Figure 5). Figure 4 View largeDownload slide A comparison with the Census data: Sample representativeness The figure further elaborates data from ISTAT (2011 census), PLANET, and ASIA. Figure 4 View largeDownload slide A comparison with the Census data: Sample representativeness The figure further elaborates data from ISTAT (2011 census), PLANET, and ASIA. Figure 5 View largeDownload slide Dynamics of firms and employment: Sample representativeness The figure further elaborates data from PLANET, ASIA and ISTAT (“Labour Force Survey”). Data on the total economy (panel b) come from ISTAT and are computed filtering the data along a sectoral composition as close as possible to that of the firms included in the sample: private nonfinancial nonprimary sectors. Figure 5 View largeDownload slide Dynamics of firms and employment: Sample representativeness The figure further elaborates data from PLANET, ASIA and ISTAT (“Labour Force Survey”). Data on the total economy (panel b) come from ISTAT and are computed filtering the data along a sectoral composition as close as possible to that of the firms included in the sample: private nonfinancial nonprimary sectors. Temporary contracts, which account for more than 10% of all contracts (Table 1), could act as a buffer for firms to adjust to a credit shock in the very short term. The average duration of temporary contracts in our sample is 9.4 months, and about two thirds of the temporary contracts end within a quarter. Table 1 Summary statistics Variable Mean SD Share in total employment (%) $$\Delta Employment$$ - Total $$-$$0.0211 0.2610 100.0 $$\quad$$Low education $$-$$0.0098 0.1480 38.7 $$\quad$$Medium education $$-$$0.0088 0.1420 48.0 $$\quad$$High education $$-$$0.0022 0.0569 13.3 $$\quad$$Under 30 $$-$$0.0025 0.1080 17.9 $$\quad$$Over 30 $$-$$0.0186 0.2090 82.1 $$\quad$$Male $$-$$0.0126 0.1690 59.9 $$\quad$$Female $$-$$0.0085 0.1470 40.1 $$\quad$$Italian $$-$$0.0179 0.2260 91.4 $$\quad$$Foreign $$-$$0.0033 0.0919 8.6 $$\Delta Employment$$ - Open-ended $$-$$0.0173 0.1890 88.7 $$\quad$$Low education $$-$$0.0080 0.1000 32.8 $$\quad$$Medium education $$-$$0.0075 0.1020 45.0 $$\quad$$High education $$-$$0.0016 0.0335 11.0 $$\quad$$Under 30 $$-$$0.0015 0.0533 15.0 $$\quad$$Over 30 $$-$$0.0158 0.1620 73.8 $$\quad$$Male $$-$$0.0102 0.1220 51.0 $$\quad$$Female $$-$$0.0071 0.0984 37.7 $$\quad$$Italian $$-$$0.0148 0.1660 80.1 $$\quad$$Foreign $$-$$0.0025 0.0596 8.6 $$\Delta Employment$$ - Temporary $$-$$0.0039 0.1660 11.3 $$\quad$$Low education $$-$$0.0016 0.1060 3.9 $$\quad$$Medium education $$-$$0.0017 0.0993 5.4 $$\quad$$High education $$-$$0.0006 0.0464 2.0 $$\quad$$Under 30 $$-$$0.0014 0.0972 5.4 $$\quad$$Over 30 $$-$$0.0025 0.1210 5.9 $$\quad$$Male $$-$$0.0019 0.1080 4.9 $$\quad$$Female $$-$$0.0019 0.1120 6.3 $$\quad$$Italian $$-$$0.0029 0.1410 9.8 $$\quad$$Foreign $$-$$0.0009 0.0717 1.5 $$\Delta Loan$$ $$-$$0.0163 0.3151 – $$CSI$$ $$-$$0.0085 0.0404 – $$EXIT$$ 0.0066 0.0811 – Variable Mean SD Share in total employment (%) $$\Delta Employment$$ - Total $$-$$0.0211 0.2610 100.0 $$\quad$$Low education $$-$$0.0098 0.1480 38.7 $$\quad$$Medium education $$-$$0.0088 0.1420 48.0 $$\quad$$High education $$-$$0.0022 0.0569 13.3 $$\quad$$Under 30 $$-$$0.0025 0.1080 17.9 $$\quad$$Over 30 $$-$$0.0186 0.2090 82.1 $$\quad$$Male $$-$$0.0126 0.1690 59.9 $$\quad$$Female $$-$$0.0085 0.1470 40.1 $$\quad$$Italian $$-$$0.0179 0.2260 91.4 $$\quad$$Foreign $$-$$0.0033 0.0919 8.6 $$\Delta Employment$$ - Open-ended $$-$$0.0173 0.1890 88.7 $$\quad$$Low education $$-$$0.0080 0.1000 32.8 $$\quad$$Medium education $$-$$0.0075 0.1020 45.0 $$\quad$$High education $$-$$0.0016 0.0335 11.0 $$\quad$$Under 30 $$-$$0.0015 0.0533 15.0 $$\quad$$Over 30 $$-$$0.0158 0.1620 73.8 $$\quad$$Male $$-$$0.0102 0.1220 51.0 $$\quad$$Female $$-$$0.0071 0.0984 37.7 $$\quad$$Italian $$-$$0.0148 0.1660 80.1 $$\quad$$Foreign $$-$$0.0025 0.0596 8.6 $$\Delta Employment$$ - Temporary $$-$$0.0039 0.1660 11.3 $$\quad$$Low education $$-$$0.0016 0.1060 3.9 $$\quad$$Medium education $$-$$0.0017 0.0993 5.4 $$\quad$$High education $$-$$0.0006 0.0464 2.0 $$\quad$$Under 30 $$-$$0.0014 0.0972 5.4 $$\quad$$Over 30 $$-$$0.0025 0.1210 5.9 $$\quad$$Male $$-$$0.0019 0.1080 4.9 $$\quad$$Female $$-$$0.0019 0.1120 6.3 $$\quad$$Italian $$-$$0.0029 0.1410 9.8 $$\quad$$Foreign $$-$$0.0009 0.0717 1.5 $$\Delta Loan$$ $$-$$0.0163 0.3151 – $$CSI$$ $$-$$0.0085 0.0404 – $$EXIT$$ 0.0066 0.0811 – The table reports the summary statistics for: (1) $$\Delta Employment$$ for different demographic characteristics, for all contracts and separately for open-ended and temporary contracts; (2) the average change in firm borrowing over two quarters ($$\Delta Loan$$); (3) the credit supply index ($$CSI$$); and (4) a binary variable identifying firms that closed their activity in a given quarter $$t$$, but were active in $$t-1$$ ($$EXIT$$). The sample is the one used in the empirical analysis, made by the universe of firms, conditional on having bank debt. The change in employment for temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. Education level are defined as low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education, based on the ISCED classification. The last column report the share of employment at the beginning of the period (end 2007) for different characteristics of contract and workers: these data are taken from the “Labour Force Survey” of the National Institute of Statistics. Table 1 Summary statistics Variable Mean SD Share in total employment (%) $$\Delta Employment$$ - Total $$-$$0.0211 0.2610 100.0 $$\quad$$Low education $$-$$0.0098 0.1480 38.7 $$\quad$$Medium education $$-$$0.0088 0.1420 48.0 $$\quad$$High education $$-$$0.0022 0.0569 13.3 $$\quad$$Under 30 $$-$$0.0025 0.1080 17.9 $$\quad$$Over 30 $$-$$0.0186 0.2090 82.1 $$\quad$$Male $$-$$0.0126 0.1690 59.9 $$\quad$$Female $$-$$0.0085 0.1470 40.1 $$\quad$$Italian $$-$$0.0179 0.2260 91.4 $$\quad$$Foreign $$-$$0.0033 0.0919 8.6 $$\Delta Employment$$ - Open-ended $$-$$0.0173 0.1890 88.7 $$\quad$$Low education $$-$$0.0080 0.1000 32.8 $$\quad$$Medium education $$-$$0.0075 0.1020 45.0 $$\quad$$High education $$-$$0.0016 0.0335 11.0 $$\quad$$Under 30 $$-$$0.0015 0.0533 15.0 $$\quad$$Over 30 $$-$$0.0158 0.1620 73.8 $$\quad$$Male $$-$$0.0102 0.1220 51.0 $$\quad$$Female $$-$$0.0071 0.0984 37.7 $$\quad$$Italian $$-$$0.0148 0.1660 80.1 $$\quad$$Foreign $$-$$0.0025 0.0596 8.6 $$\Delta Employment$$ - Temporary $$-$$0.0039 0.1660 11.3 $$\quad$$Low education $$-$$0.0016 0.1060 3.9 $$\quad$$Medium education $$-$$0.0017 0.0993 5.4 $$\quad$$High education $$-$$0.0006 0.0464 2.0 $$\quad$$Under 30 $$-$$0.0014 0.0972 5.4 $$\quad$$Over 30 $$-$$0.0025 0.1210 5.9 $$\quad$$Male $$-$$0.0019 0.1080 4.9 $$\quad$$Female $$-$$0.0019 0.1120 6.3 $$\quad$$Italian $$-$$0.0029 0.1410 9.8 $$\quad$$Foreign $$-$$0.0009 0.0717 1.5 $$\Delta Loan$$ $$-$$0.0163 0.3151 – $$CSI$$ $$-$$0.0085 0.0404 – $$EXIT$$ 0.0066 0.0811 – Variable Mean SD Share in total employment (%) $$\Delta Employment$$ - Total $$-$$0.0211 0.2610 100.0 $$\quad$$Low education $$-$$0.0098 0.1480 38.7 $$\quad$$Medium education $$-$$0.0088 0.1420 48.0 $$\quad$$High education $$-$$0.0022 0.0569 13.3 $$\quad$$Under 30 $$-$$0.0025 0.1080 17.9 $$\quad$$Over 30 $$-$$0.0186 0.2090 82.1 $$\quad$$Male $$-$$0.0126 0.1690 59.9 $$\quad$$Female $$-$$0.0085 0.1470 40.1 $$\quad$$Italian $$-$$0.0179 0.2260 91.4 $$\quad$$Foreign $$-$$0.0033 0.0919 8.6 $$\Delta Employment$$ - Open-ended $$-$$0.0173 0.1890 88.7 $$\quad$$Low education $$-$$0.0080 0.1000 32.8 $$\quad$$Medium education $$-$$0.0075 0.1020 45.0 $$\quad$$High education $$-$$0.0016 0.0335 11.0 $$\quad$$Under 30 $$-$$0.0015 0.0533 15.0 $$\quad$$Over 30 $$-$$0.0158 0.1620 73.8 $$\quad$$Male $$-$$0.0102 0.1220 51.0 $$\quad$$Female $$-$$0.0071 0.0984 37.7 $$\quad$$Italian $$-$$0.0148 0.1660 80.1 $$\quad$$Foreign $$-$$0.0025 0.0596 8.6 $$\Delta Employment$$ - Temporary $$-$$0.0039 0.1660 11.3 $$\quad$$Low education $$-$$0.0016 0.1060 3.9 $$\quad$$Medium education $$-$$0.0017 0.0993 5.4 $$\quad$$High education $$-$$0.0006 0.0464 2.0 $$\quad$$Under 30 $$-$$0.0014 0.0972 5.4 $$\quad$$Over 30 $$-$$0.0025 0.1210 5.9 $$\quad$$Male $$-$$0.0019 0.1080 4.9 $$\quad$$Female $$-$$0.0019 0.1120 6.3 $$\quad$$Italian $$-$$0.0029 0.1410 9.8 $$\quad$$Foreign $$-$$0.0009 0.0717 1.5 $$\Delta Loan$$ $$-$$0.0163 0.3151 – $$CSI$$ $$-$$0.0085 0.0404 – $$EXIT$$ 0.0066 0.0811 – The table reports the summary statistics for: (1) $$\Delta Employment$$ for different demographic characteristics, for all contracts and separately for open-ended and temporary contracts; (2) the average change in firm borrowing over two quarters ($$\Delta Loan$$); (3) the credit supply index ($$CSI$$); and (4) a binary variable identifying firms that closed their activity in a given quarter $$t$$, but were active in $$t-1$$ ($$EXIT$$). The sample is the one used in the empirical analysis, made by the universe of firms, conditional on having bank debt. The change in employment for temporary contracts includes fixed-term direct hires, project workers, temporary agency workers, trainees and apprentices, and seasonal workers. Education level are defined as low (at most compulsory education), medium (at most upper secondary education), and high (tertiary education) education, based on the ISCED classification. The last column report the share of employment at the beginning of the period (end 2007) for different characteristics of contract and workers: these data are taken from the “Labour Force Survey” of the National Institute of Statistics. Looking at the subsample of the indebted firms (i.e., those used in the empirical analysis), the average firm has 6.3 employees (the median is 2 employees); two-thirds of the firms are in the service sector. In terms of the geographical distribution, firms are roughly equally distributed across the seven provinces of Veneto, with Padua (20%) and Verona (19%) being the two most populated provinces, and Venice (the regional capital) accounting for 16% of firms. Finally, our sample mostly includes firms that borrow from one bank, while about one-third of firms were borrowing from more than one bank at the beginning of the period. The job loss for the average firm is equal to 2.1%, while credit declined by 1.6% (see Table 1) consistent with the evidence of a significant credit crunch in Italy following the Lehman’s collapse (Presbitero et al. 2014; Cingano et al. 2016).10 However, the reduction in bank credit and employment was heterogeneous, as one fourth of firms experienced a negative change in employment and credit contracted for more than half of the firms in the sample. 2. Identification Strategy 2.1 The empirical model We test for the effect of credit supply on firm employment decisions estimating the following model: \begin{align}\label{eq:eq1} \Delta Employment_{it} =\beta \Delta Loan_{it} + \delta_{i} + (\gamma_{s(i)} \times \tau_t) + (\eta_{c(i)} \times \tau_t) + (\theta_{p(i)} \times \tau_t) + \epsilon_{it} , \end{align} (1) where the changes in total employment ($$\Delta Employment_{it}$$) and in loans used by the banking system $$\Delta Loan_{it}$$ for firm $$i$$ over the quarter $$t$$, are calculated as \begin{align}\label{eq:delta} \Delta X_{it} =\frac{X_{it_1} - X_{it_0}}{0.5 \times X_{it_1} + 0.5 \times X_{it_0}}, \end{align} (2) where $$X_{t_0}$$ and $$X_{t_1}$$ are, respectively, the values of employment and bank lending at the beginning and the end of the quarter $$t$$. Variations calculated in this way are widely used because they have the advantage of being symmetric and bounded between $$-2$$ (exiters) and $$+2$$ (entrants) and they are equal to zero for firms that do not register any variation in employment or lending within the quarter (Moscarini and Postel-Vinay 2012; Haltiwanger et al. 2013; Siemer 2016).11 Since labor decisions are sticky and the real effects of a financial shock could be visible with some lag (Greenstone et al. 2014; Popov and Rocholl 2016), in the baseline specification we consider the average change in used loans over two quarters (formally, we calculate $$\Delta Loan_{it}$$ and $$\Delta Loan_{it-1}$$ and we take the average change).12 Summary statistics for these variables—for different job contracts and workers—are reported in Table 1. The estimate of $$\beta$$ gives the magnitude of the bank lending channel on employment dynamics. To assess the effect of bank lending on firm employment we face two main challenges. First, the observed amount of bank credit is the equilibrium of demand for and supply of credit. To deal with possible demand and productivity shocks we first add firm and time (quarter) effects, which allow for firm-specific time invariant demand shifters and for common global shocks occurring at a quarterly frequency. Then we saturate the model with more sophisticated (2-digit) industry$$\times$$quarter ($$\gamma_{s(i)} \times \tau_t$$) and province$$\times$$quarter ($$\theta_{p(i)} \times \tau_t$$) fixed effects, and with a set of dummies that vary across quarters and firm class size (micro, small and medium-large firms, $$\eta_{c(i)} \times \tau_t$$). The degree of granularity of these borrower fixed effects is such that our identification hinges on the assumptions that (1) firm unobserved heterogeneity that drives labor demand (i.e., managerial risk appetite) is time invariant and (2) all firms operating in the same 2-digit industry, in the same province, and in the same class size face the same demand or productivity shock in each quarter. Given that we consider the universe of firms in a relatively homogeneous region, we believe that such granular fixed effects should be sufficient to isolate time-varying unobserved demand shocks. That said, we run additional robustness test allowing for more demanding firm cluster$$\times$$time fixed effects to absorb time-varying borrower demand shocks, using industry-province-size-quarter fixed effects (see Section 5). Second, bank lending is endogenous to firms’ economic conditions and employment choices, so that standard ordinary least squares (OLS) estimates are likely to be biased.13 To isolate a credit supply shock from a lower demand for credit we build on an instrumental variable (IV) approach similar to the one proposed by Greenstone et al. (2014). We construct a time-varying firm-specific index of credit supply ($$CSI_{it}$$)—discussed in detail in the following section—and we use it as an instrument for $$\Delta Loan_{it}$$. In this way, we can measure the firm-level “aggregate” bank lending channel (Jiménez et al. 2014), which takes into account general equilibrium effects (i.e., the possibility that firms substitute for credit across banks). 2.2 Credit supply index To isolate the exogenous component of credit supply, we adopt a data-driven approach, in the spirit of Greenstone et al. (2014). Specifically, we estimate the following equation that decomposes the contribution of demand and supply factors to bank lending growth at the national level: \begin{align}\label{eq:bank} \Delta L_{bpst} = \alpha + \delta_{bt} + \gamma_{pst} + \epsilon_{bpst}, \end{align} (3) where the outcome variable $$\Delta L_{bpst}$$ is the percentage change in outstanding business loans by bank $$b$$, in province $$p$$, in sector $$s$$ at time $$t$$; specifically we observe outstanding loans for about 650 banks, 100 provinces (after excluding those located in Veneto), and the main sectors of activity (agriculture, manufacturing, construction, and private nonfinancial services);14$$\gamma_{pst}$$ is a set of province-sector-quarter fixed effects that capture the variation in the change of lending due to province-sector cycles, which can be interpreted as broadly measuring local demand; and the bank-time fixed effects $$\delta_{bt}$$ represent our parameters of interest and capture (nationwide) bank lending policies. The identification of both $$\gamma_{pst}$$ and $$\delta_{bt}$$ is guaranteed by the presence of multiple banks in each province-sector market (i.e., multiple banks exposed to the same demand) and the presence of each bank in multiple province-sector markets (i.e., multiple markets exposed to the same bank supply conditions). We then construct a time-varying firm-specific index of credit supply, aggregating the bank-specific supply shocks estimated above with the beginning-of-the-period banks’ shares at the firm level as weights. Specifically, the credit supply for the firm $$i$$ at time $$t$$ is \begin{align}\label{eq:CSI} CSI_{it} = \sum_{b} w_{bit_0} \times \hat{\delta}_{bt}, \end{align} (4) where $$\hat{\delta}_{bt}$$ are the bank-time fixed effects estimated in Equation (3) and $$w_{bit_0}$$ is the bank $$b$$ market share for firm $$i$$ at the beginning of the sample period (end-2007). By construction, $$CSI_{it}$$ captures the time-varying credit supply at the firm level and its sources of variability are the substantial heterogeneity in changes in business lending across banks and the variation in bank market shares across firms. To further convince the reader that our measure of credit supply is actually correlated with the evolution of credit conditions in Italy and with bank characteristics, we provide a set of stylized facts. First, we show that, at the nationwide level, the evolution of bank lending policies mimics quite well the growth rate of business loans; the correlation is stronger in the first part of the crisis and weaker in more recent years (Figure 6, panel a); and the latter pattern might be due to the prevalence of demand factors in the second part of the crisis as main drivers of loan growth rate. More interestingly, from a microeconomic point of view, banks applying different conditions in terms of access to credit are characterized by significant differences in loans dynamics. Specifically, for each period we divide banks into two groups, depending on whether their estimated credit supply orientation ($$\hat{\delta}_{bt}$$) was below or above the median, and we examine credit patterns for both groups: as expected, tight banks recorded more negative patterns than ease ones (Figure 6, panel b). Next, we can see that there is significant variability in credit supply across banks, with the large contraction in the supply of credit around 2009 being driven by banks with the lowest values of $$\hat{\delta}_{bt}$$ (Figure 6, panel c).15 Finally, the time pattern of our credit supply indicator is also consistent with other aggregate indicators measuring the credit supply orientation. Specifically, in panel d) of Figure 6 we plot the (inverse of) $$CSI$$ together with (1) the diffusion index from the ECB Bank Lending Survey on Italian banks,16 (2) the share of rationed firms as reported by a survey on firms maintained by the Bank of Italy, and (3) a corporate credit rationing indicator developed by Burlon et al. (2016) using bank-firm-matched data. The chart shows that the credit supply index closely follows the evolution of bank lending standards and the ones of firm financing constraints; the correlation of the $$CSI$$ with the three measures of credit constraints varies between $$0.6$$ and $$0.7$$. Figure 6 View largeDownload slide View largeDownload slide Bank lending policies and credit supply index: Descriptive statistics The time-varying nationwide bank lending policies ($$\hat{\delta}_{bt}$$) at the bank level and the credit supply index ($$CSI_{it}$$) at the firm level are obtained following the approach by Greenstone et al. (2014), as discussed in Section 2.2 (see specifically Equations (3) and (4), respectively). The credit supply index is constructed aggregating the bank-quarterly fixed effects ($$\hat{\delta}_{bt}$$) with initial banks’ market share. All charts refer to Italy. Panel (a) reports the average bank lending policy obtained averaging the bank-level $$\hat{\delta}_{bt}$$ weighted by bank market share in terms of loans. In panel (b), tight (ease) banks are those that, in each quarter, have a bank lending policy ($$\hat{\delta}_{bt}$$) below (above) the median. In panel (c) we divide banks depending their lending policies ($$\hat{\delta}_{bt}$$) and we report the evolution of $$\hat{\delta}_{bt}$$ for banks at the 25th, 50th and 75th percentiles. Panel (d) plots four indicators, all standardized to make the comparison easier: (a) the inverse of the $$CSI$$; (b) the Diffusion index, calculated from answers to question 1 (“Over the past 3 months, how have your bank’s credit standards as applied to the approval of loans or credit lines to enterprises changed?”) of the ECB Bank Lending Survey on Italian Banks (the five possible answers to questions 1 and 6 are: (1) tighten considerably, (2) tighten somewhat, (3) remain basically unchanged, (4) ease somewhat, and (5) ease considerably. The diffusion index varies between -1 and 1; it is computed as the weighted mean of answers (1)–(5), where the values attributed to each answer are 1, 0.5, 0, -0.5, and -1, and the weights are the observed frequencies. See www.ecb.int/stats/money/surveys/lend/html/index.en.html); (c) the share of rationed firms as reported by a survey on firms maintained by the Bank of Italy (INVIND): firms are considered as credit constrained if they asked banks or other financial intermediaries for more credit, and the request has been denied (even in part); and (d) a measure of corporate credit rationing: Burlon et al. (2016) identifies whether or not any bank-firm transaction is credit rationed or not through the estimation of supply and demand curves and under the assumption that the observed quantity of credit is the minimum between the demand and supplied quantities. Source: Elaboration on data drawn from the Bank of Italy SR, CR, BLS, INVIND; European Central Bank; and Burlon et al. (2016) Figure 6 View largeDownload slide View largeDownload slide Bank lending policies and credit supply index: Descriptive statistics The time-varying nationwide bank lending policies ($$\hat{\delta}_{bt}$$) at the bank level and the credit supply index ($$CSI_{it}$$) at the firm level are obtained following the approach by Greenstone et al. (2014), as discussed in Section 2.2 (see specifically Equations (3) and (4), respectively). The credit supply index is constructed aggregating the bank-quarterly fixed effects ($$\hat{\delta}_{bt}$$) with initial banks’ market share. All charts refer to Italy. Panel (a) reports the average bank lending policy obtained averaging the bank-level $$\hat{\delta}_{bt}$$ weighted by bank market share in terms of loans. In panel (b), tight (ease) banks are those that, in each quarter, have a bank lending policy ($$\hat{\delta}_{bt}$$) below (above) the median. In panel (c) we divide banks depending their lending policies ($$\hat{\delta}_{bt}$$) and we report the evolution of \hat