 |
A Cross-national Comparative Study of the Policy Effects of Referendums
| |
| | Unformatted Document Text:
cal results discussed above are obtained by estimating a simple linear regressionmodel with a dichotomous indicator for the presence of such institutions. AsFunk and Gathmann (2006) nicely argue, provided that the citizens’ preferences,which are likely to relate to the policy outcome, are different in cantons or stateswith and without institutions allowing for referendums, estimating the effect ofreferendums in such a manner yields biased estimates. Hence, this clearly pleadsfor integrating measures for the voters’ preferences. In addition, however, the em-pirical model to be estimated should also reflect more precisely the theoreticallyimplied relationship. Almost all theoretical models dealing with the policy effectsof referendums (e.g., Steunenberg, 1992; Gerber, 1999; Moser, 1996; Matsusakaand McCarty, 2001; Hug and Tsebelis, 2002; Hug, 2004; Kessler, 2005) come tothe conclusion that allowing voters to vote under certain rules directly on poli-cies should lead to outcomes more closely reflecting the voters’ wishes. Only themodels of Matsusaka and McCarty (2001), Matsusaka and McCarty (2001) andKessler (2005) suggest that under specific assumptions voters might be worse offif referendums are possible.
These theoretical implications are rarely directly and correctly tested as Mat-
susaka (2001) convincingly argues. For policy outcomes that are measured di-chotomously, the empirical models used in Gerber (1996), Gerber and Hug (1999)and Hug (2004) allow for directly testing the theoretically implied effect. Modelswith a linear specification and a simple dichotomous indicator for referendumsyield under almost all circumstances biased estimates as Matsusaka (2001), Hug(2001), and Funk and Gathmann (2006) argue, and Hug (2001) demonstrates inMonte Carlo simulations.
11
Hence, strictly speaking only the results obtained by
Gerber (1996, 1999), Gerber and Hug (1999), Hug (2001, 2004) can demonstratedirectly that in many policy areas the presence of institutions allowing for refer-endums reduces the difference between voter preferences and policy outcomes.
12
Such systematic tests, however, have not been carried out at the national
level so far. While some cross-national comparative work on referendums exists
national level without any empirical basis. Similarly doubtful about this practice are Kiewietand Szakaly (1996, 64) for the United States: ‘[t]here are many reasons to be cautious in makinginferences about the federal government on the basis of data from state and local governments.”
11
Even though Funk and Gathmann (2006) are aware of the problem, their empirical strategy
only works under very restrictive assumptions, as we will discuss below.
12
Matsusaka (2004), after estimating a simple linear regression model with a dichotomous
indicator for referendums for two time periods, namely the interwar period and the 1990s,attempts to explain the differences with extraneous information to overcome these limitations.
5
|
| | Authors: Hug, Simon. and Gilland Lutz, Karin. |
|
| |
|
|
cal results discussed above are obtained by estimating a simple linear regression
model with a dichotomous indicator for the presence of such institutions. As
Funk and Gathmann (2006) nicely argue, provided that the citizens’ preferences,
which are likely to relate to the policy outcome, are different in cantons or states
with and without institutions allowing for referendums, estimating the effect of
referendums in such a manner yields biased estimates. Hence, this clearly pleads
for integrating measures for the voters’ preferences. In addition, however, the em-
pirical model to be estimated should also reflect more precisely the theoretically
implied relationship. Almost all theoretical models dealing with the policy effects
of referendums (e.g., Steunenberg, 1992; Gerber, 1999; Moser, 1996; Matsusaka
and McCarty, 2001; Hug and Tsebelis, 2002; Hug, 2004; Kessler, 2005) come to
the conclusion that allowing voters to vote under certain rules directly on poli-
cies should lead to outcomes more closely reflecting the voters’ wishes. Only the
models of Matsusaka and McCarty (2001), Matsusaka and McCarty (2001) and
Kessler (2005) suggest that under specific assumptions voters might be worse off
if referendums are possible.
These theoretical implications are rarely directly and correctly tested as Mat-
susaka (2001) convincingly argues. For policy outcomes that are measured di-
chotomously, the empirical models used in Gerber (1996), Gerber and Hug (1999)
and Hug (2004) allow for directly testing the theoretically implied effect. Models
with a linear specification and a simple dichotomous indicator for referendums
yield under almost all circumstances biased estimates as Matsusaka (2001), Hug
(2001), and Funk and Gathmann (2006) argue, and Hug (2001) demonstrates in
Monte Carlo simulations.
11
Hence, strictly speaking only the results obtained by
Gerber (1996, 1999), Gerber and Hug (1999), Hug (2001, 2004) can demonstrate
directly that in many policy areas the presence of institutions allowing for refer-
endums reduces the difference between voter preferences and policy outcomes.
12
Such systematic tests, however, have not been carried out at the national
level so far. While some cross-national comparative work on referendums exists
national level without any empirical basis. Similarly doubtful about this practice are Kiewiet
and Szakaly (1996, 64) for the United States: ‘[t]here are many reasons to be cautious in making
inferences about the federal government on the basis of data from state and local governments.”
11
Even though Funk and Gathmann (2006) are aware of the problem, their empirical strategy
only works under very restrictive assumptions, as we will discuss below.
12
Matsusaka (2004), after estimating a simple linear regression model with a dichotomous
indicator for referendums for two time periods, namely the interwar period and the 1990s,
attempts to explain the differences with extraneous information to overcome these limitations.
5
|
|
 Convention | | All Academic Convention can solve the abstract management needs for any association's annual meeting. | | Submission - Custom fields, multiple submission types, tracks, audio visual, multiple upload formats, automatic conversion to pdf. | | Review - Peer Review, Bulk reviewer assignment, bulk emails, ranking, z-score statistics, and multiple worksheets! | | Reports - Many standard and custom reports generated while you wait. Print programs with participant indexes, event grids, and more! | | Scheduling - Flexible and convenient grid scheduling within rooms and buildings. Conflict checking and advanced filtering. | | Communication - Bulk email tools to help your administrators send reminders and responses. Use form letters, a message center, and much more! | | Management - Search tools, duplicate people management, editing tools, submission transfers, many tools to manage a variety of conference management headaches! | | Click here for more information. |
|
|
|
| |
|
|
|
