Causal Inference and Research Design

Findings, often reported in our leading journals, that genes determine our political behaviors (participation, voting, rebellion, etc.) have multiplied in recent years and have drawn much attention in the popular press. A headline like “Single Gene Predicts Conservatism,” inevitably appeals to a folk belief in simple, deterministic explanations, and perhaps even to a neo-Calvinism that absolves us of responsibility for our choices. Many in the field have expressed skepticism. In our lead article, “Candidate Genes and Political Behavior,” Evan Charney and William English pose a more fundamental difficulty: Do these studies even grasp correctly the genetics that are allegedly involved? Their answer, lucidly argued and doubtless controversial, is “no.” They offer some remedial instruction in modern genetics, denigrate the by-now routine search among “candidate genes,” and also question the results of many twin studies. Charney and English illustrate their arguments with reference to a single recent article (whose authors may wish to respond in a future issue), but their argument extends much farther. One expert geneticist, who kindly reviewed this submission for us, responded that it should “dampen enthusiasm for gene association studies” generally in political science.

Another increasingly popular line of inquiry is the “natural experiment,” where the planets allegedly align to produce an assignment to “treatment” that seems almost random. In “When Natural Assignments Are Neither Natural nor Experiments,” Jasjeet S. Sekhon and Rocío Titiunik question the validity of many such efforts and note that, for causal inference, much stronger assumptions are often required (and often not met). Sekhon and Titiunik illustrate their argument by re-examining four recent major works (including one article that appeared in this Review) that invoke either a “natural experiment” or a regression discontinuity design. In none of these cases, they argue, is assignment to the “treated” and the “untreated” groups really so random as it seems, and in one instance—a landmark article on the “personal vote”—they use new data to show that when the randomness condition is fully met, some of the reported results go away. As with the Charney and English article, Sekhon and Titiunik's critique will (we believe) spark a fruitful debate on the strengths and weaknesses of a vibrant area of research, and on the validity of our inferences in research that purports to employ natural experiments.

Policy Performance and Government Survival

Whether democracy requires equality, or equality leads inevitably to democracy, is one of the longest-running debates among students of politics: Aristotle, Tocqueville, Jefferson, and Marx (to name only a few) expressed clear—if not always empirically well-tested—views; more recently: Engerman and Sokoloff; Boix, Acemoglu and Robinson; Bartels; and Hacker and Pierson, have weighed in, and the topic has even entered electoral campaigns. Does equality lead to democracy (Tocqueville, Boix), does growing inequality hasten the revolutionary achievement of democracy (Marx) or undermine a democratic government already achieved (Jefferson), do we expect an “inverted U” relationship between inequality and democracy (Acemoglu and Robinson), or is democracy unaffected by the level of inequality among its citizens?

Any contemporary test of the link, argue John R. Freeman and Dennis P. Quinn in “The Economic Origins of Democracy Reconsidered,” must take into account a factor first emphasized by Bates and Lien and re-emphasized by Boix, namely the mobility of capital. Where capital can exit easily, democracy threatens the wealthy less and hence is more easily achieved. Global financial integration, Freeman and Quinn contend, has eased the path to democracy by making more exit options available: e.g., even large landowners can securitize their holdings on international markets and so diversify their portfolios that expropriation of land would barely threaten their wealth. Thus, they reason, democratization will be increasing in inequality among financially open autocracies but ultimately decreasing in inequality among closed autocracies, where elites feel their backs to the wall. Freeman and Quinn's empirical work, covering a panel of countries since 1945, sustains this conjecture and, paradoxically, offers support both for Boix and for Acemoglu and Robinson. In financially open autocracies, democratization becomes likelier with increasing inequality; in financially closed autocracies, we find an inverted-U relationship between inequality and democracy.

In “Democracy, War, and Wealth: Lessons from Two Centuries of Inheritance Taxation,” Kenneth Scheve and David Stasavage address a narrower but highly informative aspect of the same broad question: Do extensions of the franchise lead to more confiscatory inheritance taxes? The answer, they conclude, is pretty clearly “no.” Drawing on evidence from a large set of countries over the whole period 1816–2000, Scheve and Stasavage find almost no link between franchise extensions and inheritance taxes, but a very significant effect of mass warfare. Both autocracies and democracies often enact steep increases in inheritance taxes during, or soon after, major wars; not just to raise revenue—although inheritance taxes are easy to enforce—but apparently to achieve equality of sacrifice, when young men go into combat (often under duress) but old men do not. On the other hand, even big extensions of the suffrage show almost no link to increased inheritance taxes. Their argument seems also to hold in reverse: While not all governments lowered inheritance taxes in the half-century after World War II, most have; and they have done so only long after 1945, as the memory of “shared sacrifice” faded. If Scheve and Stasavage are correct, governments both democratic and autocratic can survive major wars only by equalizing, or credibly promising to equalize, the burdens of conflict.

In more peaceful times, elections in democracies amount to “referenda on the economy.” Poor economic performance, especially weak or negative growth in disposable income, makes incumbents' re-election less likely. That implies the existence of “electoral business cycles,” in which incumbents take care to stimulate the economy in the run-up to an election. Yet decades of research on the economically advanced democracies have found, at best. tenuous evidence of pre-election economic expansion. In “Electoral Business Cycles in OECD Countries,” Brandice Canes-Wrone and Jee-Kwang Park suggest an innovative answer to this riddle: Policy uncertainty before an election—especially when the race looks close and parties are polarized—discourages private investment in new capital and produces a “reverse business cycle,” the effects of which can attenuate or negate governments' efforts at pre-election stimulus. Examining quarterly data on ten OECD countries between 1975 and 2006, Canes-Wrone and Park find that: (a) the evidence for reverse business cycles is strong, and (b) once reverse business cycles are factored out, political (“opportunistic”) business cycles show up clearly. Thus it appears that governments try, but often fail, to stimulate their economies before elections.

While bad economic performance can undercut incumbents' chances, we also see that governments sometimes self-destruct, either making a bad situation worse or nullifying an otherwise passable performance: The governments of John Major, Gordon Brown, Jimmy Carter, and Gerhard Schröder come readily to mind. Such “death spirals” make sense, argue Torun Dewan and David P. Myatt in “Dynamic Government Performance: Honeymoons and Crises of Confidence,” if we simply note that a government's collective performance depends in large measure on the effort that individual ministers exert. When those ministers believe that their tenure will be long (unless terminated only by their own bad performance), they have strong incentives to exert maximum effort. Once ministers come to believe, however, that their government's days are numbered—it will be defeated in a coming election or on a vote of confidence—they exert less effort, thus worsening the government's performance and hastening its demise. While such a situation has, as might be expected, multiple equilibria, Dewan and Myatt show that random shocks to the government's popularity suffice to identify a unique equilibrium; and then a sequence of negative shocks can easily set off the downward spiral.

If a government's survival depends largely on whether the economy flourishes, do “unearned” inflows (foreign aid and remittances) make rulers more popular? More especially, can such inflows sustain autocratic governments that otherwise would be swept aside? Given the parallel to abundant natural resources, we might suspect this to be the case; but in “The Perils of Unearned Foreign Income: Aid, Remittances, and Government Survival,” Faisal Z. Ahmed shows—both by panel analysis and a natural experiment—that the greater the unearned foreign income an autocracy receives, the lower is its likelihood of experiencing violent protest, turnover, or regime collapse. Ahmed goes on to advance evidence that the causal mechanism is indeed the one we would suspect, namely aid- and remittance-financed governmental outlays for patronage.

It seems safe to say that the Chinese Communist autocracy is not maintained by foreign remittances, but rather by stellar economic performance. We might therefore expect, or even assume, that the party aligns incentives appropriately, advancing local leaders according to the economic success of their regions and stimulating a virtuous competition among local bosses. But a careful analysis of cadres' ranks within the post-reform Central Committee, Victor Shih and his co-authorsFootnote ¹ show in “Getting Ahead in the Communist Party: Explaining the Advancement of Central Committee Members in China,” reveals that ascent in the party hierarchy bears little connection to local economic success. As traditional Kremlinology might have had it, advancement depends more on factional loyalty, educational achievement, and supply of revenue to the central government.Footnote ² Thus the secret of Chinese economic growth, and of successful pro-growth policies, remains obscure. What incentives, if any, do local officials have, to spur (or, at least, not to inhibit) regional economic growth?

Rhetoric and Judgment

Leaders and followers might of course be motivated, in many of the cases just discussed, by powerful rhetoric: in favor of equality, growth, or the incumbent government, or opposed to wartime shirking. We too readily dismiss rhetoric as “cheap talk,” yet we sense viscerally that it can seize emotions, stiffen morale, and inspire action: Try to imagine the U.S. Civil War without the rhetoric of Lincoln, the New Deal without that of Roosevelt, the Battle of Britain without Churchill's.Footnote ³ A less familiar but supremely important example, argues Melvin L. Rogers in “The People, Rhetoric, and Affect: On the Political Force of Du Bois's The Souls of Black Folk,” is the rhetoric of W.E.B. Du Bois in his most influential work. Indeed, as Rogers shows, we can learn much about how rhetoric works from studying closely how Du Bois wields it—eliciting emotions such as sympathy and shame to shift what Rogers calls the “cognitive-affective dimension of judgment.” Du Bois's rhetorical goal in doing so was the crucial one of redefining “the people” in the American context, and more specifically to enlarge its meaning, previously defined almost unthinkingly along racial lines, to encompass African-Americans. Rogers therefore offers both an account of what Du Bois meant by “the people” (and why political theory should re-engage with that concept) and a critical understanding of Du Bois's rhetorical virtuosity.

Forum

Finally, this issue's forum revisits an article in this Review by Clifford J. Carruba, Matthew Gabel, and Charles Hankla (CGH; 2008), on the European Court of Justice (ECJ). CGH had argued, with evidence from a large dataset of the court's decisions, that the ECJ was effectively constrained less by its own reasoning and precedents than by credible threats, either of defiance by individual member states or of override by a coalition of states. Alec Stone Sweet and Thomas L. Brunell, in “The European Court of Justice, State Noncompliance, and the Politics of Override,” respectfully, but vigorously, disagree. Re-analyzing the same evidence deployed by CGH, they find neither threat credible and note that actual efforts to nullify controversial ECJ rulings came to naught.

Responding, in “Understanding the Role of the European Court of Justice in European Integration,” Carruba, Gabel, and Hankla clarify both their own argument and their points of disagreement with Stone Sweet and Brunell. Performing their own rigorous re-evaluation of the evidence, they stand by their conclusions and find little support for the other authors' interpretations.

Readers will note—indeed, both parties emphasize—that this disagreement is another skirmish in the Thirty Years' War between “intergovernmentalist” and “neofunctionalist” readings of how the European Union works. While cynics may venture to suggest that it is time for a “Peace of Westphalia” that embraces the rule of cuius regio, eius religio and leaves to local judgment the issue of the Real Presence of European Union autonomy, current events offer a new front in the conflict. As the “terrain of combat” moves from the ECJ to the European Central Bank, will the logic of neo-functionalism sustain the Euro and Greece's membership in it; or will the tug-of-war implicit in intergovernmentalism permit a Greek default and exit, causing what Haas and Schmitter called a “spill back” in the integration process? The unfolding of the present crisis will help to illuminate the unique federal, or confederal, experiment that the European Union has become.

Manuscript Formatting

Manuscripts should be no longer than 50 pages (8.5 × 11 inch paper) or 45 pages (DIN A4), including text, all tables and figures, notes, references, and appendices intended for publication. Font size must be at least 12 point for all parts of the submission, including notes and references, and all body text (including references) should be double-spaced. Include an abstract of no more than 150 words. Explanatory footnotes may be included but should not be used for simple citations. Do not use endnotes. Observe all of the further formatting instructions given on our web site. Doing so lightens the burden on reviewers, copyeditors, and compositors. Submissions that violate our guidelines on formatting or length will be rejected without review.

For submission and review purposes, you may locate tables and figures (on separate pages and only one to a page) approximately where they fall in the text, but with an in-text locator for each, in any case (e.g., [Table 3 about here]).

If your submission is accepted for publication, you may also be asked to submit high resolution digital source files of graphs, charts, or other types of figures. Following acceptance, all elements within any tables submitted (text, numerals, symbols, etc.) should be accessible for editing and reformatting to meet the journal's print specifications (e.g., they should not be included as single images not subject to reformatting). If you have any doubts about how to format the required in-text citations and/or bibliographic reference sections, please consult the latest edition of The Chicago Manual of Style (15th ed.; Chicago: University of Chicago Press, 2003) and review recent issues of the APSR.

Specific Procedures

Please follow these specific procedures for submission:

1. 1. Before submitting any manuscript to the APSR, download a PDF of the Transfer of Copyright Agreement from the Editorial Manager login page at http://www.editorialmanager.com/apsr and be sure its terms and requirements, as well as the permissions granted to authors under its provisions, are acceptable to you. A signed agreement will be required for all work published in this journal.

2. 2. When you submit (at www.editorialmanager.com/apsr), you will be invited to provide a short list of appropriate reviewers of your manuscript. Do not include on this list anyone who has already commented on the research included in your submission. Likewise, exclude any of your current or recent collaborators, institutional colleagues, mentors, students, or close friends. You may also “oppose” potential reviewers by name, as potentially biased or otherwise inappropriate, but you will be expected to provide specific reasons. The editors will refer to these lists in selecting reviewers, though there can be no guarantee that this will influence final reviewer selections.

3. 3. You will also be required to upload a minimum of two separate files.

   1. a) An “anonymous” digital file of your submission, which should not include any information that identifies the authors. Also excluded should be the names of any other collaborators in the work (including research assistants or creators of tables or figures). Likewise do not provide in-text links to any online databases used that are stored on any personal web sites or at institutions with which any of the co-authors are affiliated. Do not otherwise thank colleagues or include institution names, web addresses, or other potentially identifying information.

   2. b) A separate title page should include the full manuscript title, plus names and contact information (mailing address, telephone, fax, and e-mail address) for all credited authors, in the order their names should appear, as well as each author's academic rank and institutional affiliation. You may also include any acknowledgements or other author notes about the development of the research (e.g., previous presentations of it) as part of this separate title page. In the case of multiple authors, indicate which should receive all correspondence from the APSR. You may also choose to include a cover letter.

4. 4. Please make sure the file contains all tables, figures, appendices, and references cited in the manuscript.

5. 5. If your previous publications are cited, please do so in a way that does not make the authorship of the work being submitted to the APSR obvious. This is usually best accomplished by referring to yourself and any co-authors in the third person and including normal references to the work cited within the list of references. Your prior publications should be included in the reference section in their normal alphabetical location. Assuming that in-text references to your previous work are in the third person, you should not redact self-citations and references (possible exceptions being any work that is “forthcoming” in publication, and which may not be generally accessible to others). Manuscripts with potentially compromised anonymity may be returned, potentially delaying the review processes.

Further questions

Do not hesitate, in any cases of doubt, to consult the APSR Editorial Offices with more specific questions by telephone (310–794-1051) or by sending an e-mail to: apsr@polisci.ucla.edu

Class Use

The Comprehensive Publisher Photocopy Agreement between APSA and the Copyright Clearance Center (CCC) permits bookstores and copy centers to receive expedited clearance to copy articles from the APSR and PS in compliance with the Association's policies and applicable fees. The general fee for articles is 75 cents per copy. However, current Association policy levies no fee for the first 10 copies of a printed artide, whether in course packs or on reserve. Smaller classes that rely heavily on articles (i.e., upper-level undergraduate and graduate classes) can take advantage of this provision, and faculty ordering 10 or fewer course packs should bring it to the attention of course pack providers. APSA policy also permits free use of the electronic library reserve, with no limit on the number of students who can access the electronic reserve. Both large and small classes that rely on these articles can take advantage of this provision. The CCC's address, telephone, and fax are 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400 (voice), and (978) 750-4474 (fax). This agreement pertains only to the reproduction and distribution of APSA materials as hard copies (e.g., photocopies, microfilm, and microfiche).

The Association of American Publishers (AAP) has created a standardized form for college faculty to submit to a copy center or bookstore to request copyrighted material for course packs. The form is available through the CCC, which will handle copyright permissions.

APSA also has a separate agreement pertaining to CCC's Academic E-Reserve Service. This agreement allows electronic access for students and instructors of a designated class at a designated institution for a specified article or set of articles in electronic format. Access is by password for the duration of a class.

Please contact your librarian, the CCC, or the APSA Reprints Department for further information.

APSR Authors

If you are the author of an APSR article, you may use your article in course packs or other printed materials without payment of royalty fees and you may post it at personal or institutional web sites as long as the APSA copyright notice is included.

Other Uses of APSA-Copyrighted Materials

For any further copyright issues, please contact the APSA Reprints Department.