Articles

Practice 3.1: In Defense of the Chicken

Practice 3.1: In Defense of the Chicken

Instructions: The draft literature review below is much like the egg primacy example in Chapter 3. Take a few moments to draft a good revision, then a better revision, of this literature review. As before, all citations and claims are entirely fictitious.

Other scholars reject the egg primacy argument in favor of claims that chickens preceded eggs on the evolutionary tree. Webster (1988) argues forcefully in the Journal of Evolutionary Theory Studies (JETS) that chickens could have preceded eggs if the evolution to egg production resulted from a genetic mutation induced by strong and sustained electromagnetic pulses. The most likely source of such pulses, he argues, is “concerted and sustained attacks on Planet Earth [sic] by non-native life forms.” These life forms, he suggests, are “what are most commonly called aliens by the lay public” (385), though he refuses to speculate on their origin or their means of arrival (391). JETS is a highly respected publication with a journal impact factor of 1.7 (source: journal website), so we must take the claim – however ludicrous it seems to me – at least a little bit seriously.

On the other hand, Nicholson (1990) argues in the same journal that electromagnetic pulses are unnecessary. He claims that solar radiation – either during periods of thinned ozone during the Snowball Earth period or from a hole in the ozone layer created by the end-Permian meteorite’s entry into Earth’s atmosphere – could have resulted in similar effects without the “slightly incredible” (Nicholson 1990, 838) assumption of alien or extraterrestrial warfare. Nicholson’s evidence, however, seems mostly correlational and conjectural; he presents no direct evidence proving either of his proposed causal mechanisms.

The JETS issue where Nicholson’s article appears also contains several rebuttals and counterarguments from other researchers. Webster (1990) attempts to rebut Nicholson’s claims, though he presents no new evidence in support of his own thesis aside from an article on the website of a group supporting the search for extraterrestrial life that suggests the Earth once was closer to a pulsar star.

The following article by Watkins and Jin (1990) conducts a thought experiment in which they analyze the chain of events leading from the electromagnetic pulses to the conversion of an existing species to an egg-laying type. While the overall process involves several steps, they believe that the most important is that the electromagnetic pulses would have had to induce multiple simultaneous (or nearly simultaneous) changes in chickens’ DNA, particularly those governing reproduction. They claim that this is unlikely for three reasons. First, electromagnetic pulses are not known to cause genetic mutation; in fact, scientists have demonstrated rather conclusively that electromagnetic exposure, even to pulses, has no discernible effect on DNA mutation rates (1013). Second, these changes would have had to occur both identically and simultaneously across many chickens to produce a transition to egg-laying across the population, even if the transition took several generations. Producing mutations, any particular mutation, and simultaneous mutation are all rare events, so the joint probability of all three is “exceedingly, almost impossibly, low” (1990, 1021). Third, they argue that the genes governing reproductive processes are among the most resistant to mutation because of their central role in species preservation. The net result of all of these rare events is that the probability of mass mutation to a novel reproductive system is virtually nil (Watkins and Jin 1990, 1023). 

Other studies adopt a more anthropological approach. Ralph Loren, who is “professor emeritus at Princeton University” (Loren 1976, 25), analyzes the behavior of modern chickens to explain evolution. By studying Javanese Red chickens – the ancestor of all of today’s chicken breeds – Loren hopes to identify reasons why egg laying would be of particular reproductive and evolutionary value to chickens. He concludes that “the evolution of external reproduction techniques such as egg laying” (26) provided improved continuity of the species during a period in Javanese history in which crossing the road became substantially more dangerous. External reproduction, he notes, “allowed for surrogate mothering by other hens in the event of the untimely death of the laying hen” (28-29).  

Finally, Hill’s (2002) book builds on the study of Javanese chickens by suggesting that egg formation was an evolutionary strategy to expand the flightless species’ range in an archipelagic region such as Indonesia. Chicken eggs float in salt water, and an egg is a self-contained embryo requiring no outside nourishment. The average temperature of the Indian Ocean in the region of Indonesia is between 72 and 82 degrees Fahrenheit, which would combine with reflected solar rays to provide sufficient warmth for the egg to incubate even without a mother chicken to sit on it. Oceanic currents would sweep eggs to other land masses, where the chicks can hatch and fend for themselves (Hill 2002, 49). Provided that both male- and female- containing eggs are equally likely to disperse and survive, or that a suitable other type of bird that can mate with chickens is available in the new location, the species can thus both perpetuate and spread across the islands and onto the Asian mainland without need for flight (Hill 2002, 282).

Jackman (2004) takes issue with Hill’s (2002) concluding assumptions, arguing that the difference in weights in hen and rooster embryos would make rooster survival of the dispersion process much less likely. He proceeds to demonstrate this with the use of very sophisticated statistical tests, including a “likelihood ratio test” that gives a p value of 0.000 (Jackman 2004, 118). He also emphasizes the values of the “cutpoints,” whose values are 12.803, 19.711, and 26.557 (2004, 120). Jackman thus concludes that Hill’s assumptions are “invalid” (2004, 123) and that his conclusions “should be discarded” (2004, 124).

Progress 3.1: Practice ‘Dreadful Draft’ Literature Reviews

The draft literature reviews linked below were contributed by PhD students at a variety of universities. Unlike our previous examples, these are about actual research in political science.

Polarization in American Politics

Article 1: AMERICAN POLITICS: Polarization in American Politics

Author: Neil Lund, University of Maryland

Though most scholars agree that polarization in Congress has increased significantly since the 1970s, many question whether the polarization mirrors a trend in the general public. Dimaggio, Evans and Bryson (1996) represent one early effort to empirically examine mass polarization through an exhaustive examination of data from the General Social Survey (GSS) and National Election studies (NES). The authors test whether opinions on issues have become either more widely dispersed, or more clustered toward the poles. They find either no trend or slight depolarization on most issues, with the exception of abortion and policies toward the poor. Surprisingly, while they find that polarization in general has decreased, they also find that differences in opinions between self-identified Republicans and Democrats have actually increased.

Fiorina, Abrams and Pope (2005) find that polarization has occurred among the highly educated and engaged portion of the electorate that pays close attention to politics, but argues that the general public remains quite moderate and is increasingly turned off by the extremism in Congress and the press. Using data primarily from the Pew Research Center, Fiorina et. al. show that most Americans gravitate toward the ideological center on the issues. The authors argue that much of the much of what is referred to as “polarization” by other scholars, is really more accurately called “sorting”: ideological liberals have gravitated more toward the Democratic party while conservatives have become more consistently Republican. As a result, measures of polarization that average preferences on a range of issues show increasing distance between Republican and Democrats, but attitudes on individual issues are no more extreme than they were in the past.

Baldassari and Gelman (2008) use data from the GSS and NES, but focus on a slightly different aspect of political polarization than the one addressed by DiMaggio et. al. Rather than examining whether attitudes on specific issues have become more dispersed over time, the authors ask whether issue attitudes have become more closely correlated amongst one another (issue alignment) or with ideology or partisanship (partisan alignment). They reason that, “[f]rom a substantive point of view, it makes sense that if people align along multiple, potentially divisive issues, even if they do not take extreme positions on single issues, the end result is a polarized society [p.418]. Because they want to account for both aggregate and individual level factors, the authors use a varying intercept-varying slope regression model to estimate the trend in correlations between issues and ideology/partisanship between 1948-2004. Their results show evidence of growing partisan alignment, but offer little support for the occurrence of issue alignment. Policy preferences are more closely tied to party ID, but the general public does not hold consistent views across issues. They conclude that these findings are consistent with Fiorina’s view of partisan sorting.

Abramowitz (2011), who has been critical of Fiorina in a number of peer reviewed papers, finds that polarization on issues in the general electorate has increased, particularly since the 2004 elections. Abramowitz averages responses to several policy preferences questions on American National Election Studies from 1984-2004 to create a 15-point scale ranging from -7 (most liberal) to 7 (most conservative). He finds that, since 1984, responses have moved toward the poles: with a 13% increase in the number of responses clustered at the extreme ends of the scale. Abramowitz, like Fiorina, finds that the polarization is most pronounced among the “elite” portion of the public that pays close attention to politics, but notes that this is a large rapidly growing constituency. He concludes by noting that, rather than becoming less engaged with an increasingly polarized politics, the American public appears to be more interested and attentive than ever, and suggests political parties that are divided over the issues make politics relevant and meaningful for voters.

Hetherington and Weiler (2009) suggest that the acrimony in the current political climate runs much deeper than simple disagreements over policies. Instead,  they argue that Republicans and Democrats feel more divided today because the major disputes between the parties are rooted in fundamentally different understandings of the world. A growing number of Republicans, they argue, hold an authoritarian worldview. These individuals have a distaste for ambiguity, feel threatened by difference, and tend to value social cohesion over individual autonomy. Authoritarians “have little patience for nonauthoritarian’s hand-wringing tendency to ruminate over nuance and wash out legitimate distinctions in the swamp of relativism” (p.195). Using historical data from the NES and other sources, the authors show that attitudes on issues such as gay marriage, immigration, and national security are increasingly correlated with attitudes associated with an authoritarian worldview. As a result of competing worldviews, they argue, Democrats and Republicans disagree on a deep psychological level even when their preferences over policies aren’t that far apart.

Abramowitz, Alan I. 2010. The Disappearing Center: Engaged Citizens, Polarization, and American Democracy. New Haven: Yale University Press.

Baldassarri, Delia, and Andrew Gelman. 2008. "Partisans without Constraint: Political Polarization and Trends in American Public Opinion1." American Journal of Sociology 114.2: 408-446.

DiMaggio, Paul, John Evans, and Bethany Bryson. 1996. "Have American's social attitudes become more polarized?" American Journal of Sociology: 690-755.

Fiorina, Morris P., Samuel J. Abrams, and Jeremy C. Pope. 2005. Culture War? New York: Pearson Longman.

Hetherington, Marc J., and Jonathan D. Weiler. 2009. Authoritarianism and Polarization in American Politics. Cambridge: Cambridge University Press.

Indifenous Social Movements in Latin America

Article 2: COMPARATIVE POLITICS: Indigenous Social Movements in Latin America

Author: Neil Lund, University of Maryland

Yashar (1998, 2005) represents one of the earliest attempts to offer a comprehensive explanation for the sudden wave of indigenous mobilization in late 20th century Latin America. Yashar uses a most similar systems design to compare indigenous movements in five countries: Bolivia, Guatemala, Peru, Ecuador, and Mexico. She constructs case studies from elite interviews, fieldwork, and extensive policy analysis to pinpoint the factors that explain the divergent paths of indigenous social movements in each of the five countries. Yashar finds that indigenous communities were first moved to protest when neoliberal political and economic reforms threatened long-standing expectations of local autonomy in indigenous communities. While these threats were present throughout the region, communities formed robust national protest movements only when they were unified with other indigenous groups through dense trans-community networks and when protesters were unimpeded by government repression. Yashar’s explanation is particularly useful for, not only for explaining the precipitous rise of indigenous movements in countries like Ecuador and Bolivia, but also for explaining the puzzling failure of an indigenous movement to form in Peru. While Peruvian pueblos indigenas had faced many of the same threats as other indigenous groups in the region, their lack of unity, coupled with the threat of violent repression, hamstrung attempts to organize a unified resistance. 

Zamosc (1994) acknowledges the importance of the internal movement characteristics highlighted by Yashar, but suggests that the recent emphasis on ethnic identity among indigenous protesters is mostly a product of weakness of the traditional left. His study of the 1990 Levantamiento (uprising) by Andean indians in Ecuador finds that ethnic organizations assumed a leadership role in organizing rural protests when “ongoing crises within labor and peasant organizations” (p.56)  left a power vacuum in indigenous regions. When distressed agrarian communities found they could no longer rely on traditional left-wing organizations for representation, they turned to indigenous organizations for support. Zamosc suggests that indigenous leaders found themselves at the head of an essentially agrarian movement almost by accident.

Brysk (1996), in contrast to both Yashar and Zamosc, stresses international factors to explain the sudden rise of indigenous protest in the region. Brysk traces the rise of indigenous movements in Latin America to the global rise of international NGOs and the influence of sympathizers in the west. Indigenous movements, she argues, gained new power when indigenous groups linked their causes to emerging global movements for ethnic self-determination, civil rights, and environmental preservation. While the poverty, isolation, and history of discrimination that often characterize indigenous communities can serve to marginalize them domestically, these disadvantages are powerful sources of symbolic capital on the international stage. Brysk’s argument, though persuasive, fails to explain why indigenous movements formed in some countries in Latin America but not others.

Van Cott (2005) focuses primarily on the formation and success (or failure) of attempts by organizers to turn successful indigenous movements into viable political parties. Van Cott uses Hall’s (2003) method of “systematic process analysis” to compare the political fortunes of indigenous parties in six Latin American countries between 1980 and 2002. Her analysis finds that Indigenous parties were most likely to form in countries with decentralized political systems, easy access to the ballot, and few restrictions on forming new political parties. More importantly, Indigenous parties formed and won elections when older left-wing political parties were weak, corrupt and in decline. Indigenous candidates, often running as “outsiders” and campaigning on their opposition to political corruption, offered attractive alternatives for left-wing voters who were dissatisfied with the traditional left.

Madrid (2012), who also focuses on political parties, argues that ballot restrictions and the strength of the left alone are only weakly correlated with the success of indigenous movements that attempt to form political parties. Instead, Madrid proves that Indigenous parties are most successful when they combine ethnic appeals with populist rhetoric to attract mestizos and whites, who represent the majority of the voting population in these countries. Madrid uses regression analysis to find the correlates of district level vote shares for indigenous parties.  His results show that the most successful indigenous parties were those that managed to garner the support of at least some white and mestizo voters. This approach is problematic, however, because Madrid uses district level voting data to make an inference about individual level voting behavior. The findings would have been more convincing if Madrid had used King's (2003) Ecological Inference method, which applies a set of maximum likelihood algorithms to generate robust estimates of individual parameters using only the aggregated data. Despite these problems, Madrid’s theory helps explain why Bolivia’s indigenous led Movimiento al Socialismo (MAS) party has managed to thrive even while indigenous parties have declined elsewhere in the region. MAS, unlike indigenous parties in Colombia and Ecuador, continues to make a concerted effort to reach out beyond its base and appeal to non-indigenous poor.

Brysk, Alison. 1996. "Turning weakness into strength: the internationalization of Indian rights." Latin American Perspectives: 38-57.

Hall, Peter A. 2003. "Aligning ontology and methodology in comparative research." Comparative historical analysis in the social sciences 374.

King, Gary. 2013. A solution to the ecological inference problem: Reconstructing individual behavior from aggregate data. Princeton: Princeton University Press.

Madrid, Raúl L. 2012. The Rise of Ethnic Politics in Latin America. Cambridge: Cambridge University Press.

Van Cott, Donna Lee. 2005. From movements to parties in Latin America: the evolution of ethnic politics. Cambridge: Cambridge University Press.

Yashar, Deborah J. 1998. "Contesting citizenship: Indigenous movements and democracy in Latin America." Comparative Politics: 23-42.

Yashar, Deborah J. 2005. Contesting citizenship in Latin America: The rise of indigenous movements and the postliberal challenge. Cambridge: Cambridge University Press.

Zamosc, Leon. 1994. "Agrarian protest and the Indian movement in the Ecuadorian highlands." Latin American Research Review: 37-68.

Media and Terrorism

Article 3: INTERNATIONAL POLITICS: Media and Terrorism

Author: Kate Eugenis, University of Nevada, Las Vegas

Many scholars claim terrorists thrive in a climate of media attention.  This is due to the unique ways in which the media can influence and education a large swatch of the population.  While not the first to remark on the power and role of the news media in shaping the perceptions of the average individual and powerful decision-makers, McCombs and  Shaw (1972) are the first to prove the existence of the agenda setting ability in 1972 (McCombs & Shaw, 1972).  They were able to do this by testing a sample Chapel Hill’s residents to see if exposure to the media influenced what issues people believed to be important (McCombs and Shaw, 1972).  This is the first of many studies that demonstrates the power of media to influence what the general population believes is important.  McCombs and Shaw (1972) find that the media “exert a consider impact on voters’ judgments of what they considered the major issues of the campaign” (McCombs & Shaw, 1972, 180).  Meaning people believed the issues the media were reporting on the most were the issues that were of the greatest importance to the nation at that time regardless of any national or international occurrences of that time. 

Wanta, Golan, and Lee (2004) confirm this study in 2004 by studying if prevalence of a foreign nation in the media made that nation seem more important to the American people.  Their findings overwhelmingly support the hypothesis that more press makes people believe the country in question is vitally important to the national security interests of the United States (Wanta et al, 2004).  While Wanta, Golan, and Lee (2004) focus their attention on the relative importance of foreign countries, the same conclusions can also apply to foreign groups (such as terrorist groups) in that those which are reported on the most will also be seen as the most salient to the American people.

Writing in 1975, Brian Jenkins (1975) is perhaps one of the first to point out that the purpose of terrorism is not rampant death, but communication.  He states: “While terrorists may kill, by our standards sometimes wantonly, and while they may threaten a lot of people, the objective of terrorism is not mass murder.  Terrorists want a lot people watching and a lot of people listening, not a lot of people dead” (Jenkins, 1975, 158).  He argues that the sole purpose of terrorism is to gain an international audience and that attacks are often choreographed in order to best gain and keep the attention of the international media (Jenkins, 1975).  In order to persuade the audience of his argument he explains his points using examples of attacks, but does not rely on what has now become the traditional method of empirical hypothesis testing (Jenkins, 1975).  This is likely due to the sheer age of the article which relies more on the author’s wisdom than on evidence.

There are, however, articles that do use empirical tests in order to test these assertions.   Written in 1988 and published before the reforms in journalism that changed the way journalists are allowed to interact with terrorists on broadcast television, Tan (1988) analyzed the coverage of terrorist violence linked to the Irish Republican Army (IRA) operating out of Northern Ireland in order to determine if the mass media was actually the pawn of manipulative terrorists seeking an international audience.  Using quantitative methods and an extensive content analysis Zoe concludes “the longstanding belief that the amount of publicity terrorists receive is an important predictor of the subsequent level of terrorist violence” (Tan, 1988, 18).

Writing from Haifa, Israel (where there is, perhaps a greater awareness about the issue of terrorism), Weimann (2008) studies the effects of modern technology on the communicative abilities of the average terrorist (Weimann, 2008).  Claiming large acts of terrorism, such as the 9/11 attacks, are designed to be performances of death, he also claims terrorists are now using the new medium of the Internet in order better communicate their message to a larger audience without needing to rely on the media as a filter (Weimann, 2008).  He performs a content analysis on thousands of terrorist websites in order to better categorize the types of rhetoric and characteristics that all have in common and advocates for the protection of society while still recognizing the difficulty in this request due to the rights inherent in the First Amendment and other such laws present in western countries (Weimann, 2008).

However, there have been many articles written on the topic – some of which cast shadows of doubt on this “common knowledge.”  Scott (2001) is one such scholar.  He writes that the paranoia surrounding the control terrorists have over the media in manipulating them to achieve a specific end is just that: paranoia (Scott, 2001).  He uses empirical methods to show that once the media is saturated with terrorist coverage there is diminishing returns in which additional attacks do not gain greater coverage and should the media be covering one particular story regarding terrorism, they will not cover other attacks by other groups (Scott, 2001).  This congestion limits the incentive for more attacks (Scott, 2001).  However, the most interesting conclusion Scott comes to is: “Terrorists do battle with each other for the attention of the press.  One terrorist’s gain in media attention is another terrorist’s loss” (Scott, 2001, 225).  While this does not cast doubt on the terrorist’s need for the media, it does change how that media is used in relation to other global happenings.

Jenkins, Brian. 1975. “International Terrorism: A Balance Sheet.” Survival: Global Politics and Strategy. 17(4): 158-164.

McCombs, Maxwell E. and Donald L. Shaw. 1972. “The Agenda-Setting Function of Mass Media.” The Public Opinion Quarterly. 36(2): 176-187.

Scott, John L. 2001. “Media Congestion Limits Media Terrorism.” Defense and peace Economics. 12(3): 215-227.

Tan, Zoe C.W. 1988. “Media Publicity and Insurgent Terrorism: A Twenty-Year Balance Sheet.” International Communication Gazette. 42(3): 3-34.

Wanta, Wayne; Guy Golan and Cheolhan Lee. 2004. “Agenda Setting and International News: Media Influence on Public Perceptions of Foreign Nations.” Journalism and Mass Communication Quarterly. 81(2): 364-378.

Weimann, Gabriel. 2008. “The Psychology of Mass-Mediated Terrorism.” American behavioral Scientist. 52: 69-88.

Political Business Cycles

Article 4: COMPARATIVE POLITICAL ECONOMY: Political Business Cycles

Author: Jane Lawrence Sumner, Emory University

Edward Tufte, in 1978, was one of the first to discuss the political business cycle. He suggested that incumbent politicians are like murderers: they need not only the motive, but also the opportunity and the correct weapons. By opportunity, he meant that incumbents have to be able to control policy, and those policies have to affect voters' well-being. The weapons available to the incumbent must have five key characteristics: they must be attributable by the electorate to the correct policymakers, timeable by policymakers, targetable to deliver the best electoral benefit, palpable to the electorate, and manipulable by policymakers. Tufte tests the theory by performing a content analysis of US presidents' State of the Union addresses, focusing on social welfare and allocative policies, and finds evidence of electoral business cycles, and cites the now-famous anecdote about the Nixon Administration increasing Social Security benefits right before the 1972 election, with "Richard M. Nixon" signed clearly on the notification letter. This is a prime example of a policy that satisfies the five conditions, and particularly the attributability condition, as a letter with a large, clear signature on it makes it extremely clear to the electorate exactly to whom they owe the extra money in their pocket (palpability). Social security transfers, as well as transfers to military personnel, are also easily timeable and manipulable, meaning that not only do incumbents have the ability to manipulate them, but they can also choose when to do so, and targetable to a specific subset of voters. It is important to note about the electoral political business cycle,

In 1987, Douglas Hibbs, who has been a professor at both MIT and Harvard but is now retired, expands upon Tufte’s work and assumptions and adds more. He focuses primarily on the American case and criticizes as weak Tufte’s assumption that incumbents care first and foremost about reelection. Because of the institutional constraints of the American electoral system, however, he posits that if there is any merit to the political business cycle theory, it should be observed in incumbent American presidents, because, as he notes, "they need only to play the game once" (p. 263). Ultimately, Hibbs disproves electoral political business cycles and proves the existence of partisan political business cycles. This is the idea that in advanced democracies, left and right parties have distinct constituencies that assess macroeconomic outcomes differently, therefore the parties themselves behave differently, and observable cycles due to expansive macroeconomic policy are the result of left parties. His argument hinges largely on the tradeoff between inflation and unemployment, and he writes that the left will sacrifice inflation for full employment, whereas the right will gladly accept higher unemployment in exchange for lower inflation.

Countering Hibbs, William Clark largely agrees with Tufte and supports the theory of electoral political business cycles. Citing the Hotelling-Downs model of convergence under electoral competition, he argues that the combination of capitalism and democracy leads to a convergence of the left and right, provided new parties have to pay some cost to enter and that parties run in more than one election. Clark also question a key assumption of the theory, that that incumbents have the ability to manipulate macroeconomic outcomes, by examining how central bank independence, capital mobility, and fixed exchange rates can take policy out of the hands of politicians.     

Other institutional features beyond central bank independence and fixed exchange rates factor into the size and frequency of political business cycles as well. In his 2002 book, Macroeconomic Policies of Developed Democracies, Rob Franzese finds evidence that democracies with more frequent elections have larger cycles, measured as tax and transfer payments as a percentage of GDP. Endogenous election timing is an issue that Tufte had struggled with: do elections cause economic booms, or do economic booms cause elections? He concludes it could be either, because either incumbent governments may induce good macroeconomic conditions because they intend to call an election, or they could call an election because the economy is doing well and they want to take advantage of that. His data on the UK is not sufficient.

In a more recent (2006) article, Harvard’s Thomson Professor of Government Methods and Formal Theory, James Alt, and David Dreyer Lassen of the University of Copenhagen focus on how political polarization and institutional transparency, or the role of information, affect the political business cycle. In doing so, they address previous literature (although not the literature cited above) that suggests that political business cycles are exclusive to new or developing democracies, testing their theory on 19 advanced, industrialized OECD countries. They formally model an incumbent that can provide public goods financed by tax revenue, and who can incur deficits that have a short-term payoff, but that are costly in the long-term due to foregone future consumption. They include in the model a parameter for transparency, which is "the probability that voters observe the true level of deficits before the election" (p. 532). When the probability is one, voters have complete information. The more transparent the process, the less electioneering pays off to incumbents in terms of expected votes, because voters see the policies for the opportunistic time-bombs, essentially, that they are. Put more diplomatically, voters understand that the short term benefits have longer-term, but lagged, consequences to the economy. They also include political polarization in the model, and observe that the further apart political parties are on the ideological spectrum, the more distasteful they will find the implementation of the other party's (or other parties') preferred policies, and thus, the more they want to remain in office and the more debt they are willing to incur in order to do so.

Empirically, they construct an additive index of fiscal transparency from OECD survey data, with countries ranging from a minimum of zero (Japan) to a maximum of 11 (New Zealand). Their measure of party system polarization is the standard deviation of party-by-party mean raw scores of a twenty-point scale of parties' priorities on the tradeoff between cutting taxes and increasing public services. This variable ranges from a low of 2.2 in New Zealand to 6.6 in the United States. To detect pre- or post-election business cycles, they include dummies for the year before and the year after an election. Their dependent variable is central government fiscal surplus as a percentage of GDP. Their sample is 19 OECD countries, from 1989 to 1998. To test their hypothesis that political business cycles should be weaker or non-existent when transparency is high, because voters have more information, they interact the transparency variable with the pre-election dummy and the post-election dummy, in two separate interaction terms. In a separate statistical model, they interact polarization with both dummies, to test their second hypothesis. The interesting finding of their analysis is, as they put it, "the political budget cycle is where you can't see it" (p. 546), or, in other words, that they find statistically and economically significant evidence that low-transparency countries foster the political business cycle. They also find that political polarization can enhance the effect of transparency on the political business cycle.

In a very recent application, Brandice Canes-Wrone et al. examine the role of uncertainty in producing political business cycles, and find that policy uncertainty leads businesses and households to postpone important economic decisions, leading to what they call a reverse political business cycle. A reverse political business cycle is the opposite of the opportunistic political business cycle developed in the previous literature, and entails pre-election economic declines. Following a similar route as Alt and Lassen, they test their hypotheses specifically amongst OECD countries, although they test in ten rather than seventeen. They discuss formal models, both of their own construction as well as in the literature, and conclude that regardless of the specifics of the formalization, uncertainty induces a reduction in investment that is costly to reduce, or what they call irreversible investment. Their theory, which applies strictly to the OECD, predicts that we should observe periods of economic decline prior to elections, depending on the competitiveness of the election and the polarization of the system. However, because their theory makes predicts about only certain parts of the economy-- those dominated by irreversible investment-- their theory allows for the existence of other types of political business cycles to exist simultaneously, in other parts of the economy, such as total GDP and other sectors of the economy.

To test their theory, they collect and analyze quarterly data on nongovernmental gross fixed capital formation (GFCF). Fixed capital formation refers to things that are big, expensive, and hard to undo, hard to move, or hard to sell at anything but fire-sale prices in case of adverse policy shifts, such as machinery, heavy equipment, or construction. GFCF includes both household and business fixed capital investments, and they suggest that because businesses are likely to delay investment in the face of electoral uncertainty, households may follow their cue without knowing why the investments are being delayed. This is in a footnote and not expanded upon in any depth, but I suspect this is because pay cuts, or lack of raises, follows from the delayed business investment, which would lead to households being largely unable to make large investments. The year-over-year growth or decline in the GFCF serves as their dependent variable. They measure polarization based on manifesto scaling of a party's position on various dimensions including state involvement in the economy, the minimum being 0.01 in Germany in 1992 and the maximum being "greater than 5" in Australia in 1976. (Unfortunately, it's unclear if this measure of polarization meshes with Alt and Lassen's measure, because Alt and Lassen don't mention years in their descriptive statistics.) Rather than use the direct scores on this model, they create dummy variables for above average polarization and below average polarization. They base their measure of competitiveness on survey data, and again create dummy variables for above and below average competitiveness. They control for other economic conditions and, to address Hibbs, they control for partisanship. Ultimately, their analysis finds evidence, across a variety of model specifications, that fixed capital investment drops when partisanship, competitiveness, or both are high, and not otherwise. In addition, they find that if GFCF is removed from GDP, there is evidence that opportunistic per-election political business cycles exist.