Showing papers in "American Journal of Political Science in 2000"

Making the Most of Statistical Analyses: Improving Interpretation and Presentation

Abstract: W e show that social scientists often do not take full advantage of the information available in their statistical results and thus miss opportunities to present quantities that could shed the greatest light on their research questions. In this article we suggest an approach, built on the technique of statistical simulation, to extract the currently overlooked information and present it in a reader-friendly manner. More specifically, we show how to convert the raw results of any statistical procedure into expressions that (1) convey numerically precise estimates of the quantities of greatest substantive interest, (2) include reasonable measures of uncertainty about those estimates, and (3) require little specialized knowledge to understand. The following simple statement satisfies our criteria: “Other things being equal, an additional year of education would increase your annual income by $1,500 on average, plus or minus about $500.” Any smart high school student would understand that sentence, no matter how sophisticated the statistical model and powerful the computers used to produce it. The sentence is substantively informative because it conveys a key quantity of interest in terms the reader wants to know. At the same time, the sentence indicates how uncertain the researcher is about the estimated quantity of interest. Inferences are never certain, so any honest presentation of statistical results must include some qualifier, such as “plus or minus $500” in the present example. Making the Most of Statistical Analyses: Improving Interpretation and Presentation

Partisanship and Voting Behavior, 1952-1996

Abstract: I assess the extent of "partisan voting" in American national elections since 1952 using a series of simple probit analyses. My measure of partisan voting is sensitive both to changes in the distribution of partisanship and to changes in the electoral relevance of partisanship. I find that the impact of partisan loyalties on voting behavior has increased in each of the last six presidential elections, reaching a level in 1996 almost 80 percent higher than in 1972-and significantly higher than in any presidential election in at least 50 years. The impact of partisanship on voting behavior in congressional elections has also increased markedly, albeit more recently and to a level still well below that of the 1950s. I conclude that the conventional wisdom among scholars and commentators regarding the "decline of parties" in American politics is badly outdated. he "decline of parties" is one of the most familiar themes in popular and scholarly discourse about contemporary American politics. One influential journalist has asserted that "The most important phenomenon of American politics in the past quarter century has been the rise of independent voters" (Smith 1988, 671). Textbook writers tell their students that "Voters no longer strongly identify with one of the major parties as they once did" (Wilson and Dilulio 1995, 180) and that "the two major parties are no longer as central as they once were in tying people's everyday concerns to their choices in the political system" (Greenberg and Page 1997, 269). The most persistent academic analyst of partisan decline has argued that "For over four decades the American public has been drifting away from the two major political parties" (Wattenberg 1996, ix), while another prominent scholar has referred to a "massive decay of partisan electoral linkages" and to "the ruins of the traditional partisan regime" (Burnham 1989, 24). I shall argue here that this conventional wisdom regarding the "decline of parties" is both exaggerated and outdated. Partisan loyalties in the American public have rebounded significantly since the mid-1970s, especially among those who actually turn out to vote. Meanwhile, the impact of partisanship on voting behavior has increased markedly in recent years, both at the presidential level (where the overall impact of partisanship in 1996 was almost 80 percent greater than in 1972) and at the congressional level (where the overall impact of partisanship in 1996 was almost 60 percent greater than in 1978). Far from "partisans using their identifications less and less as a cue in voting behavior" (Wattenberg 1996, 27), my analysis suggests that "partisan loyalties had at least as much impact on voting behavior at the presidential level in the 1980s as in the 1950s" (Bartels 1992, 249)-and even more in the 1990s than in the 1980s.

Prime Suspects: The Influence of Local Television News on the Viewing Public

Abstract: Local television news is the public's primary source of public affairs information. News stories about crime dominate local news programming because they meet the demand for 'action news." The prevalence of this type of reporting has led to a crime narrative or "script" that includes two core elements: crime is violent and perpetrators of crime are non-white males. We show that this script has become an ingrained heuristic for understanding crime and race. Using a multi-method design, we assess the impact of the crime script on the viewing public. Our central finding is that exposure to the racial element of the crime script increases support for punitive approaches to crime and heightens negative attitudes about African-Americans among white, but not black, viewers. In closing, we consider the implications of our results for intergroup relations, electoral politics, and the practice of journalism. ocal television news is America's principal window on the world. Surveys of television viewing (e.g., Roper-Starch 1994), hours of daily programming (Papper and Gerhard 1999) and the actual share of the viewing audience captured by local newscasts (Hess 1991), all demonstrate the dominance of local news. In fact, people can watch live local news almost anytime-mornings, afternoons, evenings, prime time, and late night. As the amount of news time has increased, so too has competition between stations. The drive for audience ratings pushes local news organizations to favor an "action news" format. Stories about crime provide several necessary ingredients for the successful marketing of news-concrete events with powerful impact on ordinary people, drama and emotion, and, above all, attention-getting visuals. The special attraction of television to crime is reflected in the content of local television news. In a recent study of fifty-six different cities, crime was the most prominently featured subject in the local news (Klite, Bardwell, and Salzman 1997). In some cities, crime accounted for more than 75 percent of all news coverage. We argue that local news coverage of crime follows a standard script that features two distinct elements. First, crime is violent. Armed bank robberies, homicides, "home invasions," car)ackings, and gang-related activities are now staples of local news. The second element of the crime script is the presence of a particular suspect. Episodic reporting requires a regular "cast" of characters the most prominent of which is the suspect. Given the visual nature of the medium, the importance of the suspect to the script means that crime news is often accompanied by racial imagery (Campbell,

News Media Impact on the Ingredients of Presidential Evaluations: Politically Knowledgeable Citizens Are Guided by a Trusted Source

Abstract: Scholars have uniformly presumed that news media attention to a policy issue increases its impact on presidential job performance evaluations because news coverage enhances the accessibility of beliefs about the issue in citizens' memories, which automatically increases their impact on relevant judgments. The research reported here demonstrates that media coverage of an issue does indeed increase the cognitive accessibility of related beliefs, but this does not produce priming. Instead, politically knowledgeable citizens who trust the media to be accurate and informative infer that news coverage of an issue means it is an important matter for the nation, leading these people to place greater emphasis on that issue when evaluating the President. Thus, news media priming does not occur because politically naive citizens are "victims" of the architecture of their minds, but instead appears to reflect inferences made from a credible institutional source of information by sophisticated citizens. W hen television, radio, and motion pictures joined newspapers as conduits of political information during this century, scholars worried that these communication media might have powerful "hypodermic" effects on the general public's political attitudes, injecting information and opinions into people's minds (e.g., Lasswell 1927). But when this hypothesis was subjected to empirical tests, powerful persuasion by the news media appeared to be the exception rather than the rule (Klapper 1960). During the last two decades, however, it has become clear that the media do indeed shape public opinion. Not only have investigations used improved methods to document persuasion (Page and Shapiro 1992), but new media effects have been identified as well, including agenda setting and news media priming. Agenda setting occurs when extensive media attention to an issue increases its perceived national importance (McCombs and Shaw 1972). Priming occurs when media attention to an issue causes people to place special weight on it when constructing evaluations of overall presidential job performance (Iyengar et al. 1984). Our focus in this article is on the cognitive mediators of news media priming. To specify the mediators of an effect is to identify the mechanisms by which one factor affects another (Baron and Kenny 1986). In the case of priming, news media attention to an issue presumably causes a change in a

Estimating policy positions from political texts

Abstract: The analysis of policy-based party competition will not make serious progress beyond the constraints of (a) the unitary actor assumption and (b) a static approach to analyzing party competition between elections until a method is available for deriving reliable and valid time-series estimates of the policy positions of large numbers of political actors. Retrospective estimation of these positions in past party systems will require a method for estimating policy positions from political texts. Previous hand-coding content analysis schemes deal with policy emphasis rather than policy positions. We propose a new hand-coding scheme for policy positions, together with a new English language computer-coding scheme that is compatible with this. We apply both schemes to party manifestos from Britain and Ireland in 1992 and 1997 and cross validate the resulting estimates with those derived from quite independent expert surveys and with previous manifesto analyses. There is a high degree of cross validation between coding methods, including computer coding. This implies that it is indeed possible to use computer-coded content analysis to derive reliable and valid estimates of policy positions from political texts. This will allow vast volumes of text to be coded, including texts generated by individuals and other internal party actors, allowing the empirical elaboration of dynamic rather than static models of party competition that move beyond the unitary actor assumption. eriving reliable and valid estimates of the policy positions of key actors is fundamental to the analysis of political competition. Various systematic methods have been used to do this, including surveys of voters, politicians, and political scientists, and the content analysis of policy documents. Each method has advantages and disadvantages but, for both theoretical and pragmatic reasons, policy documents represent a core source of information about the policy positions of political actors. We explore various ways to extract information about policy positions from political texts. We are particularly interested in using computer-coding techniques to derive reliable and valid estimates of the policy positions of political actors. This is not mere laziness on our part, a lack of stomach for the hard graft of expert coding. If analyses of party competition are to move beyond both static models and a view of political parties as unitary actors, this requires information on the policy positions of actors inside political parties and on the development of these over time and between elections. The laborious expert "hand-coding" of text is simply not a viable method for estimating the policy positions of huge numbers of political actors, for example, all members of a legislature. Any serious attempt to operationalize a model of internal party policy competition, or of dynamic policy-based party competition or coalition government between elections, implies using computer-coding for estimating the policy positions of key political actors. We first review existing methods for estimating policy positions from political texts. These have for the most part concentrated on the expert coding of party manifestos. We then suggest ways to improve these, dealing with both expertand computer-coded content analysis. We then explore the impact of our suggestions upon estimates of party policy positions derived from British and Irish manifestos issued during the 1992 and 1997 general elections in each country, positions for which a range of

Issue Framing and Public Opinion on Government Spending

Abstract: The issue of government spending provides an interesting context for testing issue-framing effects in American public opinion. Competing partisan elites clearly portray the spending issue in different ways: Republicans tend to focus on broad, general appeals, while Democrats aim at more specific forms of programmatic expenditures. Their differing arguments undoubtedly arise because the varied issue frames generate different kinds of responses. This study uses data from the 1992 CPS National Election Study to examine the preceding hypothesis. The results from the empirical analysis show that public opinion on government spending does, in fact, vary markedly with the presentation of the issue. This framing effect is powerful enough to induce individual-level opinion change. And, framing effects arise because varying presentations of the government-spending issue activate different sets of influences on citizens' issue attitudes. These findings have broad implications concerning both the magnitude of framing effects and the explicitly political nature of the issue-framing process. his article examines issue-framing effects in American public opinion. A single social problem can be characterized and discussed in several different ways. The specific terms used to "create" a political issue out of a social problem have a strong effect on the nature and degree of popular agreement with the various sides of that issue. An obvious implication is that politicians will attempt to define, or "frame," issues in ways that maximize support for their own positions. The issue of government spending provides an interesting context for testing these ideas about issue framing. Republicans and Democrats clearly portray the use of public expenditures in different ways. Republicans focus on broad, general appeals (e.g., "government spending must be cut!"), while Democrats aim at more specific forms of programmatic outlays (e.g., "It is important to fund medical care for the elderly!"). Their differing arguments undoubtedly arise because the varied issue frames-in this case, the general presentation versus the more specific portrayal of government spending-generate different kinds of responses. In this study, I will employ public opinion data from the 1992 CPS National Election Study to examine the preceding hypothesis, along with its causes and consequences. Public opinion on government spending does, in fact, vary markedly with the presentation of the issue. And framing effects arise because varying presentations of the government-spending issue activate different sets of influences on citizens' attitudes. These findings contribute to scholarly understanding of issue framing in several ways. First, they demonstrate that significant framing effects exist outside the laboratory. The effects occur in settings that approximate the everyday world of political discourse, on an issue that is central to the predominant lines of partisan cleavage in contemporary American politics. Second, the analysis shows that framing does not merely produce different distributions of public opinion; instead, varying issue presentations can ac-

Putting Parties in Their Place: Inferring Party Left-Right Ideological Positions from Party Manifestos Data

Abstract: The left-right ideological positions of political parties play a central role in theorizing about many different aspects of democratic processes. Unfortunately, scholars are hindered in their ability to test existing theories by the limited availability of data that is comparable over time and across countries. This paper describes a simple 'vanilla' method for using manifestos data to estimate party leftright positions. It then tests this method and four existing ones by regressing a variety of accepted survey-based measures of left-right party positions on the estimates of party positions generated by the various techniques. Finally, analysis of the residuals from these regressions identifies the extent to which there are systematic sources of errors in using manifestos data to estimate party left-right positions. The vanilla method consistently produces the best estimates of party positions, and these estimates are quite good (less than one point, on average, from the estimates of other accepted approaches). Manifestos data, however, tend to locate extreme parties closer to the ideological center than do other survey-based approaches. he left-right ideological positions of political parties play a central role in theorizing many different aspects of democratic processes. Scholars employ the left-right positions of parties in theoretical arguments and empirical tests on such topics as macro-economic policymaking, legislative institutional choice, electoral competition, voting behavior, political representation, and cabinet stability. Thus, evaluation of a wide range of theoretical arguments in comparative politics requires reliable measures of the left-right positions of political parties. A practical problem facing scholars, however, is the limited availability of appropriate measures. There are two common sources of data on leftright party locations. One consists of expert surveys on specific party systems, where experts are asked to provide estimates of party left-right positions in these systems. Castles and Mair (1984) published the first such survey in a large number of countries, and results from a similar, subsequent survey are presented in Huber and Inglehart (1995).' Although these surveys are widely used to measure party positions, they suffer important limitations: they have been administered infrequently, in different formats, and only over the last fifteen years. A second source of data is the left-right positions of party supporters as reported in mass surveys such as the Eurobarometer or the World Values Survey. These surveys have been administered more frequently, but only over the last two decades and for a limited number of countries. Thus, scholars are frequently forced to limit their

Estimating Party Influence in Congressional Roll-Call Voting

Abstract: This article develops and implements a simple procedure to estimate the extent to which party influences roll-call voting in the U.S. Congress. We find strong evidence of party influence in both the House and the Senate, in virtually all congresses over the period 1871-1998. We do not find any large, systematic differences in influence between the House and Senate. Over the post-war period, party influence in the House occurs especially often on key procedural votes-the rule on a bill, motions to cut off debate, and motions to recommit. In terms of substantive issues, party influence appears most frequently on budget resolutions, tax policy, social security, social welfare policy, and the national debt limit, while it is relatively rare on moral and religious issues and civil rights, and entirely absent on issues such as gun control. On some issues, such as agriculture, public works, and nuclear energy, party influence has varied dramatically over the period we study.

Reconsidering the Environmental Determinants of White Racial Attitudes

Abstract: J. Eric Oliver is Assistant Professor of Politics and Public Affairs, Princeton University,Robertson Hall, Princeton, NJ 08544 (eoliver@wws.princeton.edu). Tali Mendelberg isAssistant Professor of Politics, Princeton University, Corwin Hall, Princeton, NJ 08544(talim@princeton.edu).We are grateful to Paul Sniderman for making the zip codes available from the 1991Race and Politics Study. This paper was greatly improved by the comments of LarryBartels, Lisa D’Ambrosio, Martin Gilens, Don Green, Taekou Lee, Karen Stenner, andthe anonymous reviewers. An earlier version of this paper was presented at the 1998Annual Meeting of the International Society of Political Psychology, Montreal, Canada.

Measuring Issue Salience

Abstract: The concept of issue salience has figured prominently in many studies of American political life. Long lines of research have taught us that both citizens and political elites may respond differently to issues that are salient to them than to those that are not. Yet analysts making such claims about elite actors face a fundamental problem that their counterparts in mass behavior do not: they cannot survey, say, members of the Supreme Court to ascertain those cases that are especially salient to the justices. Rather, scholars must rely on surrogates for issue salience-surrogates that are fraught with problems and that have led to disparate research results. Accordingly, we offer an alternative approach to measure issue salience for elite actors: the coverage the media affords to a given issue. We argue that this approach has substantial benefits over those employed in the past. Most notably, it provides a reproducible, valid, and transportable method of assessing whether the particular actors under investigation view an issue as salient or not. In making the case for our measure we focus on Supreme Court justices but we are sanguine about its applicability to other political actors

Bargaining and War

Abstract: mains possible after war starts. n his well-known book, The Causes of War (1988), Geoffrey Blainey claimed that to understand why wars started one had to understand why they stopped. There is good reason to take this claim seriously. Nearly all wars end not because the participants are incapable of further fighting but because they agree to stop, and even capitulation is often the result of concessions made by the victor (Kecskemeti 1958). Thus to explain why wars occur one must explain why states must fight before reaching an agreement. The problem of explaining the occurrence of war is therefore ambiguous: one might merely want to explain why two states could not reach agreement without fighting, or explain why their fighting lasted long enough and was severe enough to be identified as a war, or explain why some historical war occurred. And if fighting is expected to lead to agreement then fighting must be considered as part of the bargaining process and not an alternative to it. However, this is not the way most of the literature on war approaches the problem. War is instead normally assumed to be entirely the result of choices made before it begins. If it is represented as anything more than a final outcome it is usually depicted as a costly lottery, and a negotiated settlement, if considered at all, is normally represented as an alternative to war rather than something war is expected to lead to.1 The only attempt to follow Blainey's suggestion is an article by Wittman (1979) on war termination. Wittman modeled a war as a costly military contest between two states which will be won ultimately by one or the other of them. If it is to be ended instead by a negotiated settlement then both states must prefer the terms of the settlement to the expected value of continuing the war. One implication of this fact, Wittman claimed, is that changes in the relative military power of the two states will have no effect on whether they will be able to reach a negotiated settlement, since making one state more pessimistic about continuing the war will only make the other more optimistic and therefore cannot diminish the

Heterogeneity in Perceptions of National Economic Conditions

Abstract: Empirical findings concerning economic voting differ according to the level of analysis employed. A widely accepted explanation for the inconsistency between macroand microlevel evidence of economic voting is the high degree of random variation that plagues survey data. According to this explanation, aggregation purges individual-level "noise" from mass opinions on policy issues and outcomes. Building on the research of Bartels (1996), we debate the validity of this explanation by demonstrating that public evaluations of the national economy vary systematically with information, media exposure, political attitudes, personal experiences, and demographic characteristics. Furthermore, we show that these sources of subjective heterogeneity produce systematic biases when national economic evaluations are aggregated. These findings challenge a widely accepted notion that because "error" in individual-level measures (or expressions) of preferences is random, the aggregation of these individual measures ensures that they correctly represent the collective preference. mpirical findings concerning the influence of the economy on political behavior differ depending on the level of analysis employed. Using aggregate-level data, numerous studies have established a statistical relationship between economic conditions and election outcomes or executive popularity (e.g., Kramer 1971; MacKuen, Erikson, and Stimson 1992). On the basis of the macro-level evidence alone one would have little reason to doubt that the economy matters. In contrast, individual-level studies using survey data have produced only mixed evidence that voters consider economic conditions when evaluating the incumbent government and its leaders (e.g., Fiorina 1978, 1981; Kinder and Kiewiet 1979, 1981). Moreover, the strongest survey evidence of economic voting suggests that voters primarily care about national economic conditions not their personal financial situation. This finding contradicts the egocentric theory posited by early macro-level research that national economic indicators matter because they are correlated with personal financial circumstances. The micro-macro dichotomy in economic voting research reflects a similar dichotomy in research on voter sophistication and representation. Aggregate-level studies tend to indicate that voters are reasonably well informed about the economy and make forward-looking, rather sophisticated voting decisions (e.g., Alesina, Londregan, and Rosenthal 1993; Suzuki and Chappell 1996; Palmer and Whitten 1999). These studies infer that mass political decisions conform to rational expectations, or at least limited information rationality (e.g., MacKuen, Erikson, and Stimson 1992). In contrast to this macro-level image of the voter, individual-level survey research has shown that the public in general has low levels of political sophistication, lacking much knowledge of and interest in politics (e.g.,

"Old Voters, New Voters, and the personal Vote: Using Redistricting to Estimate the Incumbency Advantage"

Abstract: We provide new estimates of the size of the personal vote in U.S. House elections from 1872 to 1990. We take advantage of the "natural experiment" that attends decennial redistricting: every ten years, most incumbents are given new districts that contain a combination of old and new territory. By contrasting an incombent's vote in the new part of the district with his or her vote in the old part of the district, we can estimate the magnitude of the personal vote-the vote that the incumbent receives because he or she represented the voters in the past. Our estimates confirm prior work that shows that a large fraction of the incumbency advantage owes to the personal vote, as opposed to challenger quality. Unlike past research, we are able to estimate the relationship between district partisanship and the personal vote. We find a significant interaction which shows that incumbents develop larger personal votes in areas where they are electorally most vulnerable. here is no doubt that incumbents today have a substantial advantage over nonincumbents in U.S. House elections. The overall size of this advantage has been fairly well established, as has the fact that incumbents' electoral strength has increased dramatically over the postwar period. Measured in terms of vote share, the overall incumbency advantage grew from a modest 1-3 percentage point edge in the 1940s and early 1950s to a 7-10 percentage point edge in the 1980s and early 1990s.1 There is much less agreement about the sources of the incumbency advantage and the causes of its growth. Congressional scholars have identified a variety of factors that may be responsible, which can be grouped into three broad categories. The first, and the focus of this paper, is electoral advantage that stems from a legislator's homestyle. Members of the House are reputedly very responsive to their constituents; in turn, their constituents reward them with their votes. Mayhew (1974b), Fenno (1974), and others detail the sorts of activities that legislators do to build a widely known and respected name, including casework, frequent visits to the district, position-taking on issues, and securing federal benefits. A second source of the incumbency advantage is candidate quality. Incumbents may win because they are the best candidates running, and they do noticeably worse when they face state legislators and other experienced challengers. Finally, incumbency may simply act as a voting cue, a label which voters rely on because party has become less relevant. Assessing the importance of these different factors has significant implications for our normative evaluations of the modern congress and our understanding

A Unified Statistical Model of Conflict Onset and Escalation

Abstract: In this article I argue that much of the inconsistency in the scholarly literature on conflict escalation can be attributed to selection bias. To control this bias statistically, I use a censored probit model to evaluate hypotheses about conflict escalation over the time period of 1950-1985. I report four results. (1) Power parity and economic development are found to influence conflict nonmonotonically. (2) Although joint democracy and joint satisfaction with the status quo are found to have robust pacifying effects on the onset of conflict, the results suggest that they are unrelated to the escalation of disputes to war. (3) Allied pairs of states appear less likely to escalate their disputes. (4) Finally, the unified model suggests that it is essential for researchers interested in the escalation behavior of states to consider first how states become involved in disputes. Conflict onset and escalation appear to be related processes. any scholars have sought to isolate the causes of escalation, but discrepancies between the theoretical expectations and the empirical results persist. I attribute these discrepancies to a selection effect. Pairs of states do not become entangled in hostilities randomly. They instead select or are selected into disputes by a strategic process. Following previous empirical research focusing on selection effects (Achen 1986; Smith 1996a, 1996b, 1998, 1999; Gartner and Siverson 1996; Leeds and Davis 1997; Signorino 1999), I employ a statistical model to control for the interdependent relationship between the onset of disputes and escalation to war over the time period of 1950-1985. The results suggest that controlling for the selection effect statistically stipulates a link between the formal and empirical studies of escalation. The scholarly literature on what causes escalation is mixed. Although some suggest that joint democracy inhibits escalation (Rousseau et al. 1996), others find that joint democracy actually makes dyads more escalatory (Senese 1997). There is also some disagreement about the effect of military capabilities on conflict onset compared to escalation (Morgan 1984, 1990, 1994; Morrow 1989; Fearon 1994b; Bueno de Mesquita, Morrow, and Zorick 1997). It seems reasonable to suspect that the factors that influence conflict onset may also affect escalation directly and/or indirectly. If this is the case, it is important to consider how the factors that influence onset and escalation may be related to each other. Constructing a unified statistical model offers a first cut at modeling the process of conflict. Similar to the formal literature (Morrow 1989; Bueno de Mesquita and Lalman 1992; Fearon 1994a, 1994b; Smith 1998, 1999) that describes explicitly the selection process by which states get into and escalate disputes, I address the same issue from an empirical perspective (Smith 1996b, 1998, 1999; Signorino 1999). Modeling such selection empirically manages sources of bias and allows one to make truer inferences about the conflict generating process.

Gradations of Democracy? Empirical Tests of Alternative Conceptualizations

Abstract: measured. Should scholars use intermediate categories to measure differences between democratic and nondemocratic regimes? In a series of influential studies, Przeworski, Alvarez, Cheibub, and Limongi describe this practice as "ludicrous" and insist on dichotomous measures (Alvarez et al. 1996, 21; Przeworski et al. 1996; Przeworski and Limongi 1997, 178-179). Their position, which is shared by other prominent scholars (e.g., Linz 1975, 184-185; Huntington 1991, 11-12), is surprising. An insistence on dichotomous measures appears to neglect the advances in data collection and analysis that would allow for the more precise measurement of gradations (Bollen and Jackman 1989, 616-619). Also, their position seems insensitive to the incremental, and sometimes partial, process that characterizes many democratic transitions. Thus, dichotomous measures appear both methodologically regressive and lacking in face validity. This commitment to dichotomies in light of what seem like clear disadvantages has widened an important division among scholars about the conceptualization of democracy. Przeworski et al.'s argument, which is representative of the dichotomous view, rests on two logically independent claims-one about validity and one about reliability. Their validity claim is that democracy is first a question of kind before it is one of degree, and we cannot measure the degree of democracy across different "kinds" of regimes (Alvarez et al. 1996, 21-22).1 The solution, according to this logic, amounts to the well-known social science maxim to classify before quantifying (e.g., Sartori 1970, 1036-1040). Their reliability claim is that, even ifit made sense to measure gradations of democracy, dichotomous measures would be preferable because they contain less measurement error than do graded measures (Alvarez et al. 1996, 31). In sum, they argue that efforts to

Federalism and the Politics of Macroeconomic Policy and Performance

Abstract: Using economic data for the period between 1979 and 1995 for forty-six large federal and unitary developing nations, I analyze the impact of political federalism in the developing world on a number of measures of national economic adjustment, volatility, and crisis. The findings suggest that federalism in the ten nations where it operates has, as theoretically predicted, a negative effect on macroeconomic performance and reform. I argue that the macroeconomic and fiscal imbalances experienced by these federal nations are, in part, structurally determined by their devolved political and fiscal institutions that create incentives for subnational governments to avoid the political costs of fiscal adjustment. he political economy of stabilization and adjustment in developing nations has been a central topic of comparativists for nearly two decades. In large part this interest has been a response to the increases in international trade and capital flows, collectively known as globalization, which have increased the incentives for developing nations to discard statist models of development in favor of free market policies. With globalization, the capacity of developing nations to maintain macroeconomic stability increasingly determines their success in the search for international investment, competitiveness, and economic growth. As a result, questions about the ability of nations to adjust their economies to this reality have assumed tremendous importance. Under what conditions do governments carry through politically distasteful macroeconomic reforms? What are the key political and economic institutions that mediate a nation's insertion into and relationship with the global economy? What are the political preconditions for successful free market reforms? In addressing these questions, most research has followed in the steps of political economists such as Cameron (1984) and Gourevitch (1986), emphasizing the importance of national level institutions as mediators between the international economy and domestic politics, largely ignoring the complicated relationship between market-oriented reform efforts and the decentralization of political and fiscal power inherent in federal political systems. Likewise, the literature on the reform of the state has failed to explore the implications of federalism and subnational institutions for macroeconomic management in developing nations characterized by the devolution of political and economic power from central to provincial governments. In contrast' this research takes as its central point of theoretical departure the potentially negative consequences of federalism in the developing world for economic adjustment to the challenges of globalization.' I

Using Media-Based Data in Studies of Politics

Abstract: Political scientists have made broad use of data that are drawn from media sources. These data tend to be of two kinds: measures of the volume of media attention to an issue, or counts of events reported in the media for which there is no other systematic data source. With increasing access to full-text data archives, there will be many more opportunities for scholars to create media-based data. This paper reviews a number of issues having to do with potential bias and error in such data. Some practical solutions are described, and a media-based data series having to do with monetary policy is used to illustrate how to probe for the validity of a media-based events count. olitical scientists have often used media reports-usually print media-as the basis for data on important aspects of politics and the policy process. Many scholars drawing on data sets for measures of coups, riots, demonstrations, elections, and other political events (e.g., Polity III and the World Handbook of Political and Social Indicators) are utilizing data originally drawn from newspaper accounts. We rely extensively on "media-based event counts." With the growing availability of computerized text retrieval systems, one can anticipate even more use of such indicators in the future. It is a propitious moment to review the issues raised by these data. We use media-based counts extensively, and we probably should do so more often, especially in studies of public policy. For example, many conceptions of the policy process suggest that changing beliefs, expectations, or information of critical actors may be associated with changes in policy outcomes and outputs. Mobilizing support often depends on generating favorable media attention. Media accounts provide information on these variables over time and over political jurisdictions. I examine problems scholars confront when using media-based event counts. Scholars may easily go astray exploiting these opportunities, and using online sources does not appear to be a panacea. Many problems do not have a simple and direct solution; for others, additional empirical research may help. Here I draw broadly on research dealing with media data (see also Franzosi 1987 and Kaufman, Dykers, and Caldwell 1993). Although scholars in foreign policy and comparative politics have often examined the problems of media-based event counts, few scholars in American politics and public policy refer to that work. I shall also analyze a data series involving U.S. monetary policy, showing with a real example how scholars can probe the validity of media-based data.

Party Discipline and Measures of Partisanship

Abstract: Vote-based measures of partisanship have had a huge impact on the received wisdom about parties in the U.S. Congress. This article employs a cutpoint model to analyze how five such measures respond to changes in preference distributions and to different forms of behavior: partybased discipline and nonpartisan or undisciplined behavior. Three sets of findings are central to ongoing research about parties in legislatures. (1) The well-known and widely used party-voting score cannot discriminate between polar types of behavior. (2) All five measures encourage erroneous inferences of party discipline when only intraparty preference homogeneity may be present. (3) Of the four measures that can discriminate between partisan and nonpartisan behavior, historic congressional averages are often nominally high on a 0-100 scale, however, the averages tend to be closer to no-discipline expectations than to party-discipline expectations. Cumulatively, these findings suggest that, labels notwithstanding, vote-based measures of partisanship are ineffective instruments for detecting genuine partybased voting, party strength, and leadership support.

The Transformation of Policy Ideas

Abstract: A wave of recent scholarship, embracing both rational choice and cultural approaches, highlights the importance of collective ideas in explaining politics. What this work has not adequately addressed, however, is why collective ideas sometimes radically change. Scholars typically attribute ideational change ad hoc to creative entrepreneurial agents or to exogenous shock. But these solutions fail to account for why some entrepreneurs and not others are successful in marketing their ideas or why similar shocks lead to change in one instance but continuity in another. Explaining these variations is aided by conceptualizing ideational change as a two-stage process involving collapse and consolidation. Collapse involves societal actors reaching an agreement that the old orthodoxy is inadequate. Consolidation requires social coordination on a replacement set of ideas. Both steps involve collective ideation problems where actors may have difficulty coordinating a change in dominant ideas or have incentives to shirk facilitating efficient adaptation. The outcome of these two steps is in part endogenous to ideational structure itself. Collective ideas generate expectations, which in particular ways interact with experienced events, either inhibiting change or allowing societies to transform their own dominant notions. The approach is illustrated with examples involving collectively held causal ideas in a variety of foreign policy domains. cross a variety of literatures, scholars have taken a renewed interest in the role of collectively held ideas and beliefs in politics. Rational choice scholars have turned to collective ideas both to explain how societies make decisions under uncertainty and to resolve multiple equilibria problems (Kreps 1990; Ferejohn 1991; Garrett and Weingast 1993). A second group of scholars working more closely in a sociological/cultural framework have rediscovered and refined the explanatory power of ideas in a variety of issue areas including macroeconomic policy, trade, industrial strategy, national security, and internal security (Hall 1989; Sikkink 1991; Dobbin 1994; Katzenstein 1996a,b). But what both groups have not explained adequately is why enduring dominant ideas sometimes radically shift. What is necessary, but missing, is an explanation of discontinuous change in collective ideas.1 Explaining such change is a concern well beyond the study of politics as seen in the debates in the history and philosophy of science literature (Andersson 1994). In this article, however, I address the problem of ideational change with particular reference to policy concepts, especially foreign policy concepts held by societies we know as nation states. Why do collective ideas sometimes undergo dramatic transformation, yet at other points show basic continuity? The array of extant answers to this question is usefully examined within the agent-structure debate that has shaped economics, sociology, and political science. While each of the three main positions in this debate-agentic, structuralist, and structurationist-all provide insights, each addresses the issue of change problematically. The standard solution to their pitfalls is to invoke "exogenous shock" as the source of change-but this factor is only recognized post hoc because a change has occurred. Instead what is needed is a better-specified notion of how events relate to the causal mechanisms of the three positions, either in support of ideational continuity or as a source of change.

Severing the Electoral Connection: Shirking in the Contemporary Congress

Abstract: Studies of representation focusing on legislative shirking tend to offer an interesting tension: when the opportunity presents itself, it seems that exiting members do not change their votinq behavior (ideological shirking) but do reduce their effort level by voting less (participatory shirking). Our analysis avoids the pitfalls of prior studies by defining a simple, quasi-experimental, research design contrasting those certain that they are seeking reelection with those sure that they are departing the House for the102nd, 103rd, and 104th Congresses (1991-1996). We measure changes in ideology and participation using all roll-call votes in the last six months of election years, controlling for possible member selection effects and short-term influences, and show that legislators are disposed toward ideological and participatory shirking while selection effects are unimportant. Thus, we reconcile results seemingly at odds and suggest that policies affecting member turnover, such as term limits, can interfere with representation. A fundamental issue concerning political representation is the extent to which the necessity of reelection constrains officeholders. If incumbents plan to retire, thus freeing themselves of their electoral shackles, or if they pursue higher office and face a distinct constituency, will they behave differently? Or do members continue to behave in the same manner even in the absence of an electoral connection to the districts they represent? Answers to such questions have important implications for how we understand and model the representative process. Additionally, they have ramifications for the likely impact of public policy choices that affect member turnover rates, such as the imposition of term limits, on the political system. A primary means of drawing inferences regarding shirking-which we simply define as legislator behavior that differs from what would be observed given perfect monitoring and effective punishment by constituents-is to determine if politicians change their behavior when electoral discipline is removed. Most notably, scholars examine members' actions during their last period in office, on the grounds that those voluntarily departing can behave according to a distinct set of preferences, even if they run afoul of their districts or of their electoral coalitions. Empirical results on final period behavior are voluminous and sometimes contradictory (Bender and Lott 1996). Nevertheless, two general inferences seem to have gathered support. First, there is little evidence for shirking in terms of vote choice (ideological shirking).' While, as we will discuss shortly, there are a variety of potential explanations for such nonfindings, one intuitive and frequently offered reason is that they reflect member-specific selection effects. This implies that

Labor Organization and Electoral Participation in Industrial Democracies

Abstract: The literature on the determinants of electoral participation has paid little attention to the role of labor organization. Adopting the familiar heuristic of costs and benefits, we argue that aggregate rates of turnout will be affected strongly by the strength of the labor movement. This hypothesis is tested using cross-sectional and pooled time series data for nineteen industrial democracies and the fifty American states. The results indicate that the greater the share of workers represented by unions, the greater is the turnout. Further analysis indicates that a portion of this effect occurs indirectly through labor's ability to move the ideological position of parties appealing to lowerand middlestatus citizens farther to the left. The implications for the study of electoral politics, democratic theory, and public policy are discussed. he theoretical importance of electoral participation to democratic politics is well understood. The rate of voter participation has been considered a metric by which to judge the legitimacy of democratic institutions (e.g., Piven and Cloward, 1988), an influence on party vote shares (e.g., Nagel and McNulty, 1996), and a determinant of public policies (e.g., Hicks and Misra, 1993; for a review, see Lijphart 1997). Given the salience of turnout, it is not surprising that scholars have devoted considerable attention to the subject. Yet the literature suffers from a general failure to consider that industrial democracies are also capitalist democracies in which political competition is shaped partially by social class. The capacity of the working class, broadly construed, to compete-and thus its incentive to participate-is widely agreed to be a function of the extent to which it is organized. As Przeworski puts it, "workers can process their claims only collectively and indirectly through organizations . . . principally trade unions" (1985, 11) and, potentially, the political parties beholden to them. As we demonstrate below, labor organization is one of the principal determinants of cross-national and domestic rates of electoral participation.

Examining Latino Turnout in 1996: A Three-State, Validated Survey Approach

Abstract: As Latino populations in the United States increase, accurately characterizing their turnout is central to understanding how the post-New Deal party system will evolve. Yet we presently have little data on either their turnout or the dynamic by which such participation occurs. We estimate Latino voting rates in the 1996 presidential election by validating selfreported turnout from a post-election survey of Latinos in California, Florida, and Texas. We then use these estimates as dependent variables for multivariate models of Latino turnout. The data show that the validated Latino turnout was much lower than the aggregate turnout for the 1996 election. In addition, many of the factors that have explained aggregate voting were also significantly correlated with Latino turnout. These correlations, however, were stronger for self-reported than for validated Latino voting. Perhaps the most distinctive feature of Latino voting in 1996 was the significant and positive effect of contacting by a Latino group, which suggests that mobilization efforts may be critical to eradicating the turnout gap and incorporating Latinos into the existing party system. ith each national election during the past twenty years has come the prediction that Latino turnout would be massive and decisive. As of 1996, despite more voters and a rapidly growing population, the promise of an overwhelming Latino turnout had not been realized. There were, however, several reasons to expect that the 1996 election would witness the long-awaited dramatic increase in Hispanic voting. First, Latino leaders and officeholders spurred the Democrats to develop a major Latino outreach effort (Subervi-Velez and Cunningham 1999). Republican Latinos, although less influential than their Democratic counterparts, promoted a similar initiative (de la Garza and DeSipio 1997). Second, Latino leaders linked the anti-immigrant elements of the 1996 Welfare Reform Act to California's Proposition 187 (which was passed in 1994 and sought to deny social services to the children of illegal aliens) in hopes of mobilizing Latino voters nationwide. Third, in response to provisions in the new welfare reform, Latino immigrants began naturalizing at a record pace, greatly increasing the potential numbers of Latino voters (Freeman et al. 1998). Finally, motor-voter provisions and Latino registration drives helped raise the number of Latino registered voters to 6.6 million, almost 1.5 million more than 1992 (Southwest Voter Registration and Education Project 1997). The 1996 election is thus exceptionally interesting to students of Latino voting. In addition to a growing population, increased registration, and a favorable political context for Latino participation, concerted efforts to mobilize Latinos may have increased turnout. Here we examine the influence of mobilization and other political and demographic variables on turnout in 1996. Our data come from a three-state, post-election survey of Latinos. We find that validated Latino turnout was much lower than esti-

Dynamic modeling for persistent event-count time series

Abstract: cycles in international relations. Scholars use event-count models to analyze a wide variety of political science data. For example, the large sets of data on cooperative and conflictual international events often use event-count models (Huang, Kim, and Wu 1992; Sayrs 1992; Volgy and Imwalle 1995). These data have also been analyzed using time-series methods, particularly when transformed into indices of conflict and cooperation that consider the severity of each event and are no longer purely event counts (Goldstein 1991; Goldstein and Freeman 1990; Schneider, Widmer, and Ruloff 1993; Ward and Rajmaira 1992). Other event counts commonly analyzed in international relations include the number of wars (Benoit 1996; Mansfield 1992), militarized interstate disputes (Gowa 1998; Pollins 1996; Senese 1997), or other incidents of conflict and cooperation between states (BrophyBaerman and Conybeare 1994; Eyerman and Hart 1996; Kinsella 1995; Kinsella and Tillema 1995; O'Brien 1996; Fordham 1998a, 1998b; Remmer 1998). Scholars in American politics have also employed event-count models for time-series data to analyze presidential activity (Brace and Hinckley 1993), Federal Reserve decisions (Krause 1994), federal contract awards (Mayer 1995), executive orders (Krause and Cohen 1997), and Supreme Court decisions (Caldeira and Zorn 1998). The same issues confront the smaller number of comparativists using such data (Moore, Lindstrom, and O'Regan 1996).

What Is a Language Community

Abstract: Theories of nationalism, democracy, regional assertiveness, and civil war have relied on vague and unspecified notions of linguistic heterogeneity, based upon estimates of the "mother tongues" of a population. Under conditions of mother-tongue diversity, the criterion for a language community that requires structural proximity of languages is equally problematical, and here linguistics gives us little theoretical guidance or empirical data. Another criterion of language diversity measures the probability that two random people in a country will share a language. Empirical problems (on getting standards of "knowing" a language) and theoretical ones (concerning whether the ability to communicate is sufficient for an indicator of cultural homogeneity) beset these measures. In light of these problems, the paper specifies several new measures that might be used for coding a language community. Data collected from six post-Soviet republics illustrate the potential usefulness of these measures. thnic heterogeneity is often portrayed as a powerful source of democratic instability, regional assertiveness, and civil war. In his classic essay on primordial conflict, Geertz (1973) sees it as a source of chronic tension in the postcolonial states after World War II. Dahl (1971) sees it as a serious constraint to the success of democracy. Rabushka and Shepsle (1972) model ethnic heterogeneity such that it leads in equilibrium to the breakdown of democratic regimes. Connor (1994) equates ethnic heterogeneity with higher probability for civil war. But not all studies link heterogeneity with unhappy outcomes. Lijphart (1977) for one showed the possibility for democracy (of the nonmajoritarian sort) under conditions of cultural pluralism. The debate on the influence of ethnic heterogeneity on political outcomes has remained vibrant for nearly forty years. Yet, a major problem in testing such claims is lack of consensus on how to measure either the degree of ethnic heterogeneity within a polity or the degree of difference between any two ethnic groups in that polity. It may seem obvious that the Hirschman-Herfindahl concentration index used for political parties would be ideal for an index of ethnic dispersion-and thus all you need to know is the percentage of each ethnic group of an entire population. However, members of ethnic groups are not out there to be counted as are votes for political parties. In fact, the clear identification of ethnic groups as entities is often the result of their mobilization due to political instability and regional conflicts of interest. But if ethnic mobilization becomes the criterion for ethnic groupness, there is a problem, as the value of the independent variable becomes dependent on the value of the dependent variable. The methodological task is to develop valid indicators of ethnic heterogeneity that are independent of our observations on the dependent variables whose values are being explained. One way out of this problem is to use language as a proxy for ethnicity. The percentage of the population whose historic language is Welsh, or Catalan, or Luo therefore substitutes for a more subjective accounting of the percentage of Welsh, Catalans, or Luos in Britain, Spain, or Kenya. Use of language data in this way allows us both to measure the degree of group

Institutional Reform and Decision-Making Efficiency in the European Union

Abstract: tional reform. ince the 1970s, legislative activity of the European Union (EU) has expanded greatly in both scale and scope.' The number of legislative acts adopted by the EU per year increased from less than 300 in the mid-1970s to more than 500 in the mid-1980s. At the same time, the EU gradually extended its competencies to issue areas not explicitly covered by the Treaties of Rome, such as consumer protection, research and development, and the environment. This continuous expansion of EU legislative activity has been accompanied by periodic changes of the EU's institutional framework. The Single European Act (1987) introduced qualified majority voting for a number of policy areas and provided the European Parliament with the ability to influence legislative outcomes. The Treaty on European Union (1993) extended the use of qualified majority voting and strengthened the role of the Parliament in the legislative process. In this article, we analyze whether institutional reform has enabled the EU to deal efficiently with an expanding legislative agenda. A common theme in the literature is that the efficiency of the EU decision-making process has deteriorated considerably as EU legislative activity has increased over the past two decades. A number of studies suggest that the Council is unable to cope with the Commission's legislative output (Scharpf 1988; Sbragia 1993; Dehousse 1995). Other studies point to a dilution in the substantive content of EU legislation (Van den Bos 1994; Heritier 1996; Scharpf 1997). The most common suggestion is that the EU decisionmaking process has become inordinately slow, suffering from an excessive

The Development of Party Identification in Post-Soviet Societies

Abstract: of how the parties are arrayed along axes of political conflict and where they fit in the political spectrum. T he development of mass partisanship and party identification has long been a focus of social science inquiry. Yet, virtually all of the studies offering explanations and theories of partisan formation and change among the mass public are based on evidence derived from societies with fully developed party systems (Campbell et al. 1960; Lipset and Rokkan 1967; Converse 1969; Sartori 1976). The newly emerging democracies of the post-Soviet societies, however, offer a unique opportunity for testing theories describing the development of mass political parties and the rise of party identification. The earliest scholarly expectations regarding how post-Soviet citizens would react to the appearance of multiple political parties reflected the socialization theories prevalent in the literature. The pioneering work on party development emphasized a slow process of family and peer socialization leading to a party identification that presumably strengthens over the life cycle (Greenstein 1965; Jennings and Niemi 1974). Moreover, the traditional literature suggests that party differences and enduring party alignment arise gradually out of the important underlying group cleavages that characterize all societies (Lipset and Rokkan 1967). Thus far a prevalent view among scholars who study the most significant post-Soviet country, namely Russia, is that political parties have failed to attract popular loyalty. Indeed, many authors assert that the current post-Soviet political parties are ephemeral, impotent, and undifferentiated (see, for example, Reddaway 1994; Rutland 1994; Ordeshook 1995). In their groundbreaking work, How Russia Votes, White, Rose, and McAllister contend that "more than three quarters of the electorate lack any part identification" and that "the majority are to a significant degree anti-party" (1997, 135, 137). Rose (1994) has also concluded that the rise of political parties and party identification in Russia will take a good deal of time because post-Communist societies lack a civic culture and because Russians are

Media Framing and the Dynamics of Racial Policy Preferences

Abstract: … the symbols of revolt are not drawn like musty costumes from a cultural closet and arrayed before the public. Nor are meanings unrolled out of whole cloth. The costumes of revolt are woven from a blend of inherited and invented fibers into collective action frames…. – Sidney Tarrow (1998) After at least a decade of leftward drift, racial policy preferences reached a liberal peak in the mid-1960s. But, as we have seen, the era of liberalism was not sustainable in the long run. After a reversal and a long slow but steady march to the right, the public reached a conservative zenith in policy preferences on racial matters around 1980. That conservatism, too, was short-lived, though few observers have noted this fact: opposition to government activism on race eroded steadily throughout the 1980s. And again, in the early 1990s, the public hit a liberal high, but – in what is by now a familiar pattern – support for liberalism has been waning since then. Why? In search of an explanation, ought we to open the historian's toolkit and point to particular dates, identify “critical moments” in our history where we, like a sailboat, changed tack? Ought we to search through the records of speeches by influential politicians and societal leaders – for surely the eloquence attained in the pursuit of racial equality, with its rich biblical imagery, is among the pinnacles of rhetorical splendor in the history of the republic – to discover how a particular presidential candidate or movement activist stated something with a new and profound kind of power and persuaded plain folk to change their minds?