Quality & Quantity 34: 367–378, 2000.
© 2000 Kluwer Academic Publishers. Printed in the Netherlands.
Measuring Electoral Control
HENK VAN DER KOLK
Faculty of Public Administration, University of Twente, PO Box 217, 7500 AE Enschede, The
Netherlands, e-mail: email@example.com
Abstract. If voters do not pay attention to what representatives do, representatives are not stimulated
to be responsive. Therefore, electoral control, the extent to which voters base their vote on the
behaviour of representatives is, at least potentially, an important variable in the explanation of the
behaviour of representatives. Moreover, electoral control seems to be an important variable from a
normative point of view. In this article four ways to measure the minimal level of local electoral
control using the outcomes of local elections (electoral statistics) are presented. The general idea
behind the measures is fairly simple. If local elections are completely determined by nonlocal factors,
then the losses and gains of local divisions of national parties from one local election to another are
‘identical’ across municipalities. A deviation from this pattern can be interpreted as an indication of
the minimal level of local electoral control in a speciﬁc municipality. The measures are externally
validated using data from a survey among council members.
Key words: (sub-national) elections, voting behavior, measures.
According to William Riker, who adheres to the so called ‘liberal view of voting’,
‘the function of voting is to control ofﬁcials, and no more’.
In his opinion, ‘popular
election and a limited tenure of ofﬁce are sufﬁcient conditions for democracy’.
Implicitly, Riker assumes that in the end voters will throw the rascals out: ‘the
liberal hope is that ofﬁcials will be restrained from (unpopular policies that ofﬁcials
belief necessary) out of fear of the next election’. However, if voters do not use
information about the promises and legislative behaviour of their representatives,
the rascals can do whatever they want. For if voters do not pay attention to what
politicians do, there will be no relationship between the behaviour of representat-
ives and their chances of re-election. And if representatives know that voters are not
taking their performance into account in passing electoral verdicts, representatives
are no longer stimulated to be ‘responsive’. Therefore, electoral control, the extent
to which voters base their vote on the behaviour of representatives to be judged in
the election at hand is, at least potentially, an important variable in the explanation
of the behaviour of these representatives. Moreover, electoral control seems to be
This article is based upon parts of the Dutch thesis written by the author and published in 1997.
The paper was presented at the Annual meeting of Dutch political scientists in 1998.