The remarkably slow growth of women’s representation in most legislatures continues to pose an interesting puzzle for research. Why do women continue to be substantially underrepresented although women gained the suffrage in most democratic countries nearly a century ago? The subject is generally interesting for what it can tell us about how electoral mechanisms affect all forms of group representation. We know that proportional representation facilitates the representation of women in legislatures, but its effect varies. To accelerate the representation of women, gender quotas began to be adopted by some political parties a generation ago, and on a national scale beginning in some countries in the 1990s. In the article that opens this issue of the Quarterly, Leslie A. Schwindt-Bayer offers a comprehensive analysis of the effect of these quotas across 26 countries. She shows that details of the design of the quota decisively influence its effect. Critical factors are the size of the quota, whether the electoral law requires women candidates to be placed in winnable constituencies or winnable positions on party ballots, and the mechanisms for enforcement. Only quota size is independently important but placement requirements and enforcement mechanisms greatly increase the effect of quota size. Descriptive representation is especially important for strengthening the legitimacy of the legislature in new democracies. To strengthen their legitimacy and to attract women voters, some parties in countries having quota laws adopt even stricter quotas than the national law requires.
The second article in this issue examines the perennial question of agenda formation from a novel point of view, that of the factors which determine U.S. Senators’ selective attention to issues. Jonathan Woon measures issue attention by bill sponsorship. He expects that Senators select bills to sponsor by estimating the prospect of their passage or at least their favorable echo in their constituency. His data consist of all public bills introduced by senators who served complete terms between 1989 and 2000, divided into nine sets of issues. He finds that with the exception of health policy, committee membership is the most important factor determining issue attention, that leadership position in the committee increases issue attention, and that constituency demand is the second most important factor. The significance of Woon’s finding is that it attests to the importance of committees in an institution usually regarded as decentralized, as one whose members are unspecialized and highly individualistic. It suggests that senators, like members of the U.S. House, use their committee positions to advance the policy preferences of their constituents.
One of the advantages of comparative research across the U.S. state legislatures is that it provides opportunities for testing hypotheses about the structure of the U.S. legislative system. Thad Kousser and Justin H. Phillips examine executive-legislative relations in the budget process in 47 of the 50 states, with data over a 16-year period from 1988 to 2004. Instead of the spatial and setter models of legislative-executive interactions employed in the study of the U.S. Congress, the authors employ a “staring match” model of the relationship between governors and legislatures. They hypothesize that at the state level, each side has an incentive to reach an agreement within a limited time. The authors therefore expect that patience and endurance on each side overwhelm all other factors that might explain the predominance of the legislature or the executive in determining the size of the enacted budget. The authors demonstrate that professionalized legislatures with long sessions are likely to be more patient than their “citizen” counterparts. They find strong evidence that the length of legislative sessions, rather than any other aspect of legislative professionalization, explains relative executive-legislative influence. Their analysis suggests that the models used to examine legislative-executive relations at the congressional level, where continuing resolutions can endlessly delay budget enactments, may be inappropriate when applied to the U.S. states. It is a reminder that models employed in empirical analysis must fit the institutional context to which they are applied.
In our previous issue we began a special series of articles that deal with the problem of measuring the policy preferences of legislators. As we noted then, this problem becomes ever clearer in the context of comparative legislative research. Eduardo Alemán and his colleagues use data on cosponsorship to estimate legislators’ policy preferences in the Argentine Chamber of Deputies and the U.S. House of Representatives. They recognize that the decision on whether or not to cosponsor a bill differs from the decision on whether to vote for or against it. They take account of this difference by first creating agreement matrices that measure the ratio of bills cosponsored by every pair of legislators to the total number of bills sponsored by each of them. The authors derive ideal point estimates for each legislator using a principal components analysis of the data contained in these cosponsorship matrices. They also recognize that the highly disciplined parties that exist in the Argentine Chamber differ from the parties in the U.S. House and that for this and other reasons the institutional constraints on members differ between the two legislatures. Finally, they note that a majority of bills in each chamber have no cosponsors. Within these limits, the authors demonstrate a very high correlation between ideal point estimates derived from roll-call votes and cosponsorship data. Their analysis covers a 30-year period for the U.S. Congress and a 14-year period for the Argentine Chamber for that substantial body of bills that do have cosponsors. Since roll-call data are scarce for many of the world’s legislatures, the demonstration of the validity of an alternative measure of legislators’ policy preferences is a valuable contribution to comparative legislative research. Alternative measures of legislators’ ideal points also reveal differences in how legislators reveal their preferences in different activities and in different institutional settings.
The article by Sebastian M. Saiegh offers another alternative to roll-call votes to measure the policy preferences of legislators. He uses responses to elite surveys to place respondents in a multidimensional policy space. After discussing the shortcomings of roll-call votes as sources of data and the advantages of elite surveys, Saiegh offers an analysis of the location of legislators as well as of parties and chief executives in a common ideological space in nine Latin American countries. He shows that survey-based estimates of legislators’ policy preferences correlated well with estimates based on roll-call votes where these are available. The advantage of survey data is that it permits comparisons of policy preferences, identified as locations in ideological space, across political institutions regardless of the availability of roll-call data.
We will publish further articles on the problem of measuring the policy preferences of legislators in subsequent issues of the Quarterly.
Return to February 2009 Titles
Return to LSQ home page