517

brought to bear. I believe a serious discussion of their properties could havegreatly enriched the book.

But this critique does not take away from the great achievement inherent inthe major findings: the authors’ development of the fair-division proceduresmentioned above. The volume is valuable because of what must be termedthe head-turning original contributions of the authors.

JOE OPPENHEIMER, Government and Politics, University of Maryland, Col-lege Park, MD. 20742 U.S.A.; joppenhe@bss2.umd.edu.

Notes

1. It is interesting to note the choice criteria utilized in these algorithms: minimax. Thisshould be quite clear from the well known and simple Johnny splits, Cathy chooses, cakedivision algorithm. Johnny’s division of the cake into 2 equal pieces comes from hisattempt to insure that he can get at least half. This can lead to inefficiencies. If the cake isnot homogeneous, and the preferences are not identical, Johnny (who say, likes chocolate)could slice the cake differently. He could leave one (smaller) piece with more chocolate,and another large piece for Cathy (who, let’s assume does not prefer chocolate). Bothcould be expected to prefer that outcome.

2. So, for example, the algorithm which is developed by the authors for the n-person fairdivision of a single object requires that the object be cut into 2n�2 + 1 pieces. For n = 2this is 5 pieces, but for n = 7 it is 33 pieces. Hence, one can imagine that the step by stepdescription of a real life application of the procedure gets burdensome. For n = 4, it takesabout 2 pages of text and 6 diagrams.

3. Envy and non-envy are not unknown motivations in politics. The great levelling ideology ofthe Chinese revolution used a promise of non-envy (and perhaps even partial achievementof such results among poor peasants) to generate envious criticisms to buttress the powerand status of its elites (Chang, 1991).

References

Chang, J. (1991). Wild swans: Three daughters of China. New York: Anchor-Doubleday.Elster, J. and Roemer, J.E. (Eds.). (1991). Interpersonal comparisons of well being. Cambridge:

Cambridge University Press.

Public Choice 93: 517–522, 1997.

John M. Carey, Term limits and legislative representation. Cambridge, UK:Cambridge University Press, 1996. xiii + 216 pages. $59.95.

John M. Carey’s Term limits and legislative representation is one of themore unusual contributions to the growing body of scholarship analyzing the

518

effects of term limitations on the behavior of legislators. This work is largelytheoretical because the constitutional amendments imposing term limits onstate lawmakers in the United States are only now becoming effective. Topresent some much-needed empirical data, Professor Carey has studied CostaRica, which, in 1949, adopted a rotation requirement for all deputies in thelegislative assembly. Although former deputies can run for reelection after afour-year cooling off period, the provision effectively acts as a one four-yearterm limit for lawmakers. Since the adoption of term limits, 87 percent ofdeputies have served only one term; nearly half of those who have served twoterms sat out for at least two interim elections; and no one has served morethan three terms. Thus, Costa Rica provides an example of a political systemwhere the direct electoral connection between voters and legislators has beensevered.

Ideally, to isolate the effects of term limits on legislators, Carey would havecompared the behavior of Costa Rican deputies before and after the adoptionof term limits. The paucity of information concerning the Assembly beforethe ratification of the Constitution of 1949 makes this comparison impossible.Instead, Carey compares Costa Rica’s political situation to that of Venezuela,a country with similar political institutions but without legislative term lim-its. The countries also share similar histories, demographic characteristics,economies, and social structures. In the end, Carey’s comparative analysisleaves readers with more puzzles than solutions to the questions he raises, butin identifying these puzzles, he makes a valuable contribution.

Carey concludes that Costa Rican legislators are political careerists, thusundermining the myth of the citizen-legislator that suffuses virtually all of therhetoric from the U.S. term limits movement (Garrett, 1996). The shape ofpolitical careers in a system with term limits is, however, distinctive – at leastthe part of a career that follows one term of service in the Assembly. Deputiesarrive in the Costa Rican Assembly already highly experienced in politics,with 70 percent of national lawmakers having held prior elected office, atleast three nonelected political or party appointments, or both. Unlike U.S.or Venezuelan legislators, who then usually spend the rest of their politicallives serving in the federal legislature, Costa Rican deputies seek to extendtheir careers after service in the Assembly by obtaining political appointmentsfrom the incoming president.

Carey first demonstrates the careerist orientation of legislators by provid-ing results of his interviews with current deputies: for example, 30 of 33interviewees indicated that their expected “compensation” for faithful partyservice was a future political appointment. Carey then tests the accuracy ofthese expectations by determining how many ex-deputies actually receivedpost-Assembly appointments to executive ministries, the foreign service, or

519

directorships of influential autonomous institutions (like banks, utilities, andthe social security administration). The results, which underestimate the rateof appointments because records are incomplete, indicate that, from 1974 to1990, 64 percent of retiring deputies whose party or coalition won the nextpresidential election received political appointments.

Carey’s findings expose the puzzle of who really controls these appoint-ments, and thus to whom must career-minded deputies respond. His answer isthat ambitious deputies must align their political fortunes with the man whowins the intraparty competition for the presidential nomination. The presi-dent plays the central role in appointing literally thousands of governmentjobs that offer the promise of a longer political career. Costa Rican politicalparties apparently play little or no role in this process, so they are hamperedin maintaining party discipline in the Assembly, which is characterized byless party cohesiveness than Carey finds in Venezuela.

For Venezuelan politicians, unconstrained by legislative term limits, a polit-ical career is built through long service in the legislature (by the end of 1993,the average legislator had 11 years of experience). These lawmakers mustrespond to the entity that controls their ability to remain in the legislature:the national executive party committee which entirely determines who willappear on the congressional ballot. Not surprisingly, then, breaches in partydiscipline are virtually nonexistent in Venezuela. In Costa Rica, parties donot exert this kind of influence over political careers, presidents do; accord-ingly, Carey explains, deputies are less concerned with following the partyline when they vote, and more concerned with backing the right presidentialcandidate.

Although his hypothesis could explain the difference in party cohesiveness,it seems problematic for at least two reasons. First, because of the lack of data,Carey could not test whether Costa Rican deputies who supported the winningpresidential candidate early in the nomination process fared significantlybetter in the appointments process than their less savvy copartisans. Withoutthese data, Carey’s specific conclusions about the importance of presidentsrelative to parties are supported only by anecdotal evidence from a publicizedintraparty dispute in 1985.

Second, this portion of Carey’s book suffers from a larger problem in hisstudy. He provides little explanation of the wider political culture in CostaRica, saying nothing about campaign finance laws, the process used to selectparty leaders, the structure of the party caucus, the existence of informal partyorganizations, and the like. In this section, for example, a detailed descriptionof the appointments process might help him explain convincingly (in theabsence of any empirical evidence) why party leaders do not play a largerrole in appointing former deputies to executive posts. Contrary to Carey’s

520

hypothesis, one could tell a story in which a political party could exert agreat deal of influence by offering its help to a newly-elected president whomust quickly fill thousands of positions. Perhaps Carey might respond that,although the appointments process is a formidable task handled by numerousaides and influenced by party officials, the president personally focuses onthe fortunes of the fewer than three dozen deputies who could have helpedhim in his campaign. The president will block the appointment of a deputywho supported one of his opponents regardless of the party’s wishes.

One would wonder further, however, why the party does not insist thatthe president deal evenhandedly with deputies who supported the party’spolicies. The party appears to be in a stronger position than the president, whocan serve only one four-year term and thus influence political opportunitiesonly for this short period. Carey does note that, unlike the U.S. Senate, theAssembly lacks the power to confirm presidential appointments to prestigiousposts. Nevertheless, party leaders exert significant, and perhaps decisive,power over the fate of a president’s policy agenda in the Assembly; surely,they have various informal mechanisms to ensure that they play a role inshaping political careers. The point is that we do not know which account iscloser to the truth. Without more information about the dynamic interactionamong the party, the executive branch and the Assembly, Carey’s conclusionthat presidents control the political fortunes of former deputies and that thissituation reduces party discipline is plausible, but not compelling.

Notwithstanding this lingering puzzle, it seems clear that term-limiteddeputies rely on some partisan force – either party leadership or the newly-elected president – for the continuation of their careers. Because both theseentities are pursuing electoral goals, Carey’s observation that the electoralconnection felt by deputies has not disappeared, but has merely changedform, is persuasive. The next question is how a more indirect electoral con-nection shapes legislative behavior. Carey’s answer, that deputies still spend agreat deal of their time engaged in constituent service and pork barrel politics,is an empirical addition to the literature studying the effect of term limits onthe amount of government spending and its character (e.g., Elhauge, 1997;Aka et al., 1995).

Carey finds that term-limited deputies spent “more time providing particu-laristic goods than on any other task” (p. 104), where particularism is definedas providing goods and services exclusively to a specific set of recipients,even though the collective cost may exceed the collective benefits. Deputiesdo not spend their time and energy in these activities to ensure their reelection;instead, they are acting as agents of their parties. The parties work to sendpork to voters to achieve victory for their candidates in the next presidentialand legislative elections. As Carey nicely puts it, a system in which political

521

parties rely on deputies to seek and distribute government benefits “puts apersonal face on partisan representation” (p. 115).

To prove that parties benefit from particularism and use control over careersto get legislators to hand pork out (often literally delivering checks personallyto lucky constituents), Carey measures the impact of budget line items, calledpartidas especificas or PEs, on a party’s electoral success and on deputies’future appointments. PEs are appropriations for very small projects, costingfrom a few hundred dollars to a few thousand, that make up approximately 2percent of the annual budget. Again, for a variety of reasons, the informationCarey uses is incomplete, and the conclusions he draws tentative.

Perhaps the real problem with Carey’s analysis of particularism’s relation-ship to political careers lies in the inability to measure the kind of legislativebehavior that would have the most effect on electoral outcomes. Given thesmall variance in line item amounts allocated to each majority party deputy,legislators must distinguish themselves not in their ability to obtain PEs, butin the home styles they use to distribute the government largesse (Fenno,1978). They will be rewarded on the basis of the quality of their constituentservice, their ability to garner favorable publicity for the party when theydistribute checks or fund projects, and similar activities. Proficiency in theseimportant legislative functions (and in ensuring that the party receives credit)could be at least as important to a future political career as backing the rightpresidential candidate early in the nomination process.

Reasoning that an effective home style should result in an electoral payoff,Carey attempts to measure a deputy’s performance in these areas indirectlyby measuring electoral returns in his district (bailiwick). Carey tests whethera gain in a district in party performance during a presidential election isrelated to the deputy’s success in the appointments process; his results areinconclusive. Carey suggests that his results indicate that another variableis crucial for political careers – early support for the winning presidentialcandidate – but again the evidence to support this hypothesis is not available.

Carey ends with a study of the “last term effect” in the U.S. Congress,testing whether U.S. representatives who aspire to state-wide offices altertheir behavior to attract the support of the target offices’ constituencies. Heconcludes that aspiring representatives with relatively extreme ideologiestend to alter their voting behavior in their last terms, and that this changereveals their attempt to attract more moderate voters. He theorizes that theshift represents a strategy by relatively extreme politicians to appeal to astatewide constituency, because, as Madison argues in Federalist #10, thegreater heterogeneity of larger electoral districts requires candidates to adoptmore moderate positions to succeed in elections.

522

Carey states that his results demonstrate that aspiring members of Congressshow increased sensitivity to party interests in their last term. In fact, his studyshows only that progressively-ambitious politicians are concerned with theviews not of political parties, but of the group of voters who can elect themto their next jobs. He acknowledges as much when he clarifies that he is notclaiming that “parties as organizations systematically wield sanctions” overaspiring lawmakers (p. 176). Carey does not attempt in this book to answerthe difficult questions of whether, given the political culture in this country,U.S. political parties could control the behavior of term-limited legislators,and whether parties (or perhaps interest groups) could develop a system thatwould reward loyal members of Congress with long careers in other politicaljobs when their terms in the legislature ended.

Carey’s unique contribution to the term limits scholarship is his sustainedstudy of the operation of legislative term limits in Costa Rica, but his analysiswould have been richer had he located it in the larger political culture of Cos-ta Rica. His conclusions regarding careerism and particularism neverthelessraise puzzles that further studies should address. Specifically, the experienceof Costa Rica indicates that features of the U.S. political system decried bysupporters of term limits, such as excessive or inefficient government spend-ing, might well persist through different means. Because of the limitationsin the data about political institutions in other countries, such studies shouldfocus not on trans-national comparisons, but on U.S. states’ experience withterm limits, particularly states like California that had a tradition of profes-sionalism before the imposition of term limits. The primary objective shouldbe to identify the shape of the political opportunity structure after the adoptionof term limits because, as Professor Carey persuasively argues, the entitiesthat control politicians’ future career prospects are likely to influence stronglytheir behavior in the legislature.

ELIZABETH GARRETT, Law, University of Chicago, Chicago, IL 60637. Iappreciate the financial support of the Russell Baker Scholars Fund at theLaw School and comments by Bernard Meltzer and Tracey Meares.

References

Aka, A. et al. (1995). Is there a “culture of spending” in Congress? Mimeo.Elhauge, E. (1997). Are term limits undemocratic? University of Chicago Law Review 64.Fenno, R. (1978). Home style: House Members in their districts. Boston: Little, Brown.Garrett, E. (1996). Term limitations and the myth of the citizen-legislator. Cornell Law Review

81: 623–697.