Discrete event history analysis (EHA) is the analytic tool of choice for many scholars of policy diffusion across American states. Unfortunately, the policy diffusion literature largely ignores several important specification issues for EHA models: duration dependence, choice of functional form, and the computation of standard errors corrected for temporal and spatial dependence. We use data from Berry and Berry's (1990) seminal study of state lottery diffusion to demonstrate ways to deal properly with these issues.

The Great Recession of late 2007 through 2009 had profound negative economic impacts on the U.S. states, with 49 states experiencing revenue decreases in their 2009 budgets representing more than $67.2 billion USD. Also during this period, states enacted a record number of laws related to immigrants residing in their states. We make use of data from the National Conference of State Legislatures (NCSL) to examine punitive immigration policy enactment from 2005 to 2012 and conduct a state comparative study using cross-sectional time-series analysis to examine the potential ways in which the economic recession and changing demographics in the states have impacted punitive state immigration policy making. We hypothesize that although anti-immigrant anxieties are driven in part by economic insecurity, they are also impacted by the presence of a large or growing proportion of racialized immigrants. We find that increases in state Hispanic populations and state economic stressors associated with the recession have both led to a greater number of enacted punitive state immigration policies. In addition, we find that changes in the non-Hispanic white populations in the states are also impacting the expression of anti-immigrant attitudes in state policy during this period.

The number and variety of state policies regulating abortion each year is increasing. Opponents of abortion adopted a strategy of “legal but inaccessible” that has resulted in the passage of more than 700 state laws since the early 1990s. Despite being a very active area of policy making, we lack a coherent explanation for the proliferation of abortion policy. Scholars studying different policies at discrete moments in time have come to conflicting conclusions about how well theories of morality policy and representation explain abortion policy. Using an original dataset comprised of a near-universe of pro- and anti-abortion rights policy from 1973 to 2013, I establish the ways in which partisan control of the government and the moral preferences of constituents shape state policy. I find that anti-abortion rights policies are well explained by both theories but that pro-abortion rights policies are not well explained as a morality policy or with descriptive representation. In addition, I show the heterogeneous effect of representation across anti-abortion rights policies; Democratic women and governors decrease the probability of only certain anti-abortion rights policies.

Policy learning has been one of the primary mechanisms by which policy innovations are assumed to diffuse from one jurisdiction to another. Recent research suggests, however, that learning is more than simply observing policy adoption in other jurisdictions and must also include an assessment of the outcomes or effectiveness of those policies. This article argues that an assessment of the ability to implement an innovation is also likely to be a component of the learning process. Specifically, it argues that potential adopters are likely to ask not only, “Was the policy effective in other states that adopted it?” but also, “Can we make the policy work for us?” We test hypotheses drawn from this general argument in a directed dyad analysis of renewable portfolio standards in the American states between 1997 and 2009. Results indicate that shared implementation environments among jurisdictions help determine the impact that information about policy effectiveness has on adoption decisions. These findings suggest that implementation concerns may be an important part of the policy learning process.

This article explains governors' decisions to support or oppose Medicaid expansions offered under the 2010 Patient Protection and Affordable Care Act. We theorize that governors' decisions to oppose the funding should depend on both political demands and the level of need in the state, though politics and need are often in tension. We find that governors' partisanship and the composition of the legislature have substantively meaningful effects on governors' decisions, but the level of need in the state exerts little effect on governors' decisions. This suggests that for high profile, highly politicized issues such as the Affordable Care Act, political considerations outweigh the needs of citizens and state economic conditions in gubernatorial decision making.

The author improves on the measurement of U.S. state public opinion by (1) applying previous methods used only on cross-sectional data to create dynamic measures of state public opinion and (2) providing a systematic comparison of the performance of the various methodological approaches on these dynamic measures. The author shows that scholars can use multilevel regression, imputation, and poststratification (MRP) coupled with a simple moving average to measure state public opinion over time. Compared to aggregation, the MRP approach has less error and is more reliable, particularly for the less populated states. The author shows the applicability of the MRP approach by measuring and validating state partisanship and state ideology over time. The validated measures are available for public use. Armed with a method to measure state public opinion over time, scholars can begin to more fully understand the dynamic relationship between public opinion and policy in the U.S. states.

Scholars have suggested that interest groups affect the diffusion of innovations across states by creating a network of information between the states that aids in the spread of policy ideas. Still, the unique role that interest groups play in policy diffusion networks is not fully understood, in large part because the current methodology for studying diffusion cannot parse out interest group influence. We address this problem by analyzing the actual text of legislation, moving away from binary adoption to a more nuanced measure of policy similarity. This allows us to distinguish whether states emulate other states or interest group model legislation. We use text similarity scores in a social network analysis to explore whether early-adopting states or interest groups are more central to the network. We apply this analytical framework to two policies—abortion insurance coverage restrictions and self-defense statutes. Based on this analysis, we find that a fundamentally different picture of policy diffusion networks begins to emerge—one where interest group model legislation plays a central role in the diffusion of innovations.

The passage of a restrictive immigration law in Arizona in 2010 rekindled an old debate in the United States on immigration policy and the role of federalism. Despite periodic constitutional controversies, scholars of federalism and U.S. state politics have not adequately explained variation in state-level policy making on immigration. The authors explore pressures leading to state immigration policy innovation and adoption in the United States. The article evaluates factors leading to the introduction and adoption of two types of policies: those dictating the cultural and economic incorporation of immigrants and those attempting to control their flow and settlement. Factors such as fiscal federalism, ethnic contact, and ethnic threat generate incentives for states to pass such laws. The authors compiled a comprehensive data set of state immigration laws from the past decade to explain how factors commonly associated with national immigration policy development—economic conditions, rates of immigration, demographics, party control, and political institutions—influence state-level immigration policy activity.

The surveys of Senate races from 1988 to 1992 conducted by the American National Election Study include a number of issue questions that can be aggregated to measure state public opinion. A simple-to-compute coefficient is used to judge the aggregated reliability of these measures. A comparison of these state public opinion measures on specific issues to more general indicators, such as state ideology and partisanship, demonstrates their usefulness in obtaining a richer depiction of public preferences for studies of state politics.

Elite interviews have long been a staple of state politics research. Improved information technology facilitates the use of elite interviews, but also underscores the need for attending to their design, operationalization, and analysis. This essay provides a framework for developing, conducting, and interpreting elite interviews, and suggests means to enhance the validity and reliability of findings by considering instrumentation, sampling, data collection, and transcript analysis.

Is public policy in states that allow initiatives and referenda more responsive to public opinion than in states that do not? Political science theory provides conflicting answers to this question. On one hand, these direct democracy mechanisms give citizens a direct voice in public policy, which may directly and indirectly shape policy to their wishes. On the other hand, formal theories of collective choice call into question the ability of direct democracy to produce policy that reflects the underlying distribution of mass preferences. This study tests these competing hypotheses by assessing the impact of public opinion on abortion policy using a new survey-based measure of state-level abortion attitudes (Brace et al. 2002). The empirical findings indicate that states with initiatives and referenda are more responsive to public opinion on abortion policy.

A well-established scholarly tradition links lower voting costs with higher turnout. Whereas previous research emphasized the costs imposed by requiring voter registration, our research assesses postregistration costs and state policies that can make it easier for registered citizens to vote. These policies include mailing each registrant a sample ballot and information about the location of his or her polling place, providing a longer voting day, and requiring firms to give their employees time off to vote. Using the 2000 Voter Supplement to the Current Population Survey, we find that all but the last of these provisions enhance turnout, especially by the young and the less educated. Compared to a state that does none of these things, the estimated turnout of high school dropouts is nearly 11 percentage points higher in a state with these “best practices”; their effect on young registrants is nearly 10 points. Because African American and Latino registrants are disproportionately younger and less educated, they would benefit disproportionately from universal adoption of such postregistration laws. We estimate that if every state adopted these best practices, overall turnout of those registered would increase approximately three percentage points.

U.S. state politics researchers often analyze data with observations grouped into clusters. This structure commonly produces unmodeled correlation within clusters, leading to downward bias in the standard errors of regression coefficients. Estimating robust cluster standard errors (RCSE) is a common approach to correcting this bias. However, despite their frequent use, recent work indicates that RCSE can also be biased downward. Here the author provides evidence of that bias and offers a potential solution. Through Monte Carlo simulation of an ordinary least squares (OLS) regression model, the author compares conventional standard error (OLS-SE) and RCSE performance to that of a bootstrap method that resamples clusters of observations (BCSE). The author shows that both OLS-SE and RCSE are biased downward, with OLS-SE being the most biased. In contrast, BCSE are not biased and consistently outperform the other two methods. The author concludes with three replications from recent work and offers recommendations to researchers.

Recent elections have witnessed substantial debate regarding the degree to which state governments facilitate access to the polls. Despite this newfound interest, however, many of the major reforms aimed at increasing voting convenience (i.e., early voting and no-excuse absentee voting) were implemented over the past four decades. Although numerous studies examine their consequences (on turnout, the composition of the electorate, and/or electoral outcomes), we know significantly less about the factors leading to the initial adoption of these policies. We attempt to provide insights into such motivations using event history analysis to identify the impact of political and demographic considerations, as well as diffusion mechanisms, on which states opted for easier ballot access. We find that adoption responded to some factors signaling the necessity of greater voting convenience in the state, and that partisanship influenced the enactment of early voting but not no-excuse absentee voting procedures.

Does state political ideology change over time? Brace et al. (2004, 537) say no, based on their analysis of three longitudinal measures of state citizen ideology: Berry et al.‘s (1998) indicator that relies on election results and congressional roll call votes, and two indicators that Brace et al. construct from ideological self-placement items, one using GSS and ANES survey results, and the other employing surveys conducted by CBS/New York Times. The authors imply that the ideological stability they detect precludes the possibility that state citizen ideology influences state policy. However, this implication stems from Brace et al.‘s definition of meaningful ideological change as differences in the relative ideology of states over time rather than absolute changes in ideology within states. We contend that this argument is both logically and methodologically flawed. Brace et al. maintain that their CBS/New York Times and GSS/ANES indicators are valid measures of state citizen ideology, but that the Berry et al. indicator is not. To assess this claim, it is crucial to distinguish between ideological self-identification (or symbolic ideology) and policy mood (or operational ideology). We find that the Berry et al. measure is a valid indicator of policy mood, but that it is invalid as a measure of self-identification. In contrast, the CBS/New York Times and GSS/ANES measures are invalid as indicators of policy mood, and while they are valid indicators of self-identified ideology, they are highly unreliable. Although a measure of self-identified ideology can be useful for answering some research questions, we contend that an indicator of policy mood is more appropriate when studying the impact of public opinion on public policy, and we reiterate our confidence in using the Berry et al. (1998) measure for that purpose.

Scholars have begun to move beyond the dichotomous dependent variable—indicating whether a state adopts a policy or not in a given year—usually employed in event history analysis. In particular, they have devoted increasing attention to the components of policies that states adopt. I discuss a variety of estimators that have been employed to analyze the adoption and modification of policies with multiple components, including various forms of event history analysis, OLS, and event count models. With various modifications, the researcher can estimate models that treat each component as distinct, pool these models to leverage commonalities across components, or treat the components as identical parts of the same process. Each of these has its strengths and may be appropriate in certain circumstances. Nonetheless, in the majority of cases, some version of event history analysis for multiple or repeat failures is likely to be preferred. The different approaches are illustrated by studying state adoption of various obesity-related policies.

American chief executives are popularly believed to play a crucial leadership role in the legislative process, but this role is not well understood by scholars. This article develops a model of executive leadership success in the legislative arena based on personal, institutional, political, and economic factors and tests that model with data on the legislative priorities of the 50 state governors in 1993-94. Institutional features of the executive, which have been the focus of much scholarship, are found to have little effect on a governor's legislative success. On the other hand, the nature of the legislature is quite important here, with more professional legislatures actually encouraging executive leadership. A governor's personal characteristics, political party, and a state's economy also influence a governor's legislative success. These findings point to the importance of examining chief executives in the context in which they work if we are to understand their role as legislative leader.

Google Insights for Search provides a new and rich data source for political scientists, which may be particularly useful for state politics scholars. We outline the prior uses of Google Insights for Search in social and health sciences, explain the data-generating process, and test for the first time the validity of this data for state politics research. Our empirical test of validity shows that Google searches for ballot measures' names and topics in state one week before the 2008 Presidential election correlate with actual participation on those ballot measures. This demonstrates that the more Internet searches there were for a ballot measure, the less likely voters were to rolloff (not answering the question), and establishes the construct validity for this data for one important topic in state politics research. We also outline the limitations to this data source.

The 1990s saw some of the most dramatic changes in the American social welfare system in recent decades at both the national and state levels. In particular, states were granted, and took advantage of, much wider latitude in deciding who is eligible to receive welfare benefits. To what extent did the composition of a state's electorate influence the restrictiveness of the welfare eligibility requirements it adopted at this time? We find that in states where lower-class voter turnout was comparable to that of the upper class, lawmakers were less likely to pass restrictive welfare eligibility rules. We also find that electorates in states with restrictive voter registration laws are much more likely to be biased toward upper-class turnout. Thus, lower-class voter mobilization can affect the ability of the disadvantaged to achieve policies consistent with their interests, but state voter registration laws pose a substantial barrier to such mobilization.

Immigration remains a powerful and recurrent feature of American politics. Of the issues related to immigration, controversy over government policy for controlling illegal immigration occupies a central position in the debate. One increasingly important and prevalent type of control policy that has received little scholarly attention is worksite employment eligibility enforcement, otherwise known as E-Verify Laws. In the present article, we analyze variation in E-Verify policy adoption across the U.S. states, approaching the topic from multiple theoretical perspectives and testing several hypotheses pertaining to policy enactment. Our analysis points to the critical role of proportionate change in a state's immigrant population, as well as the political activity of immigrant-employing industries, in leading to policy adoption. Despite the use of multiple objective indicators, we fail to find strong evidence supporting the hypothesis that economic distress within a state increases its likelihood of enacting E-Verify legislation. Overall, our analysis contributes to an underdeveloped area of immigration policy research and sheds light on an important contemporary immigration issue, while drawing broader conclusions concerning the factors influencing the emergence of anti-immigration policies more generally.