|
Some Pro Term Limit Arguments
| quote: |
Incumbency Advantage
All of the arguments for term limits are premised on the costs of incumbency advantage. There is little disagreement about the importance of incumbency advantage and its increase over the past several decades. Gary King and Andrew Gelman (1991), for example, find that incumbency increases a representative's expected vote share by 11 percent. There are many possible explanations for why incumbents are advantaged but none are entirely satisfactory. Gerrymandering (redistricting) has undoubtedly benefited some representatives (perhaps at the expense of others) but incumbent senators also have an advantage over challengers and this cannot be explained by the redistricting hypothesis (Reed and Schansberg 1992). Political Action Committees (PACs) give nearly eight times as much money to incumbents than challengers. Partly as a result, incumbents spend more than two and half times as much as challengers (Maisel 1990, Abramowitz 1991). But do PACs give money to incumbents because they know the incumbent is more likely to win or are incumbents more likely to win because PACs give them more money? Some degree of co-causation is undoubtedly involved but its precise nature is difficult to identify.
Although the causes of incumbency advantage are uncertain there is a large body of evidence which shows that incumbency advantage is important and has been increasing over time. Incumbents are more likely to be reelected than challengers, they have higher margins of victory, and they participate in more unopposed elections than do challengers (Mayhew 1974, Maisel 1990, Reed and Schansberg 1992).
What are the costs of incumbency advantage and consequent lengthy terms? Term limit proponents allege incumbency advantage creates politicians who shirk their responsibilities and deviate from the wishes of the voters. I refer to this as the shirking-deviation theory. The hypothesis is that incumbency advantage serves as a relaxation of the election constraint: As reelection becomes more likely, politicians can afford to ignore the wishes of the voters. A second and related hypothesis states that even a superior challenger cannot replace an inferior incumbent. This hypothesis does not require that the incumbent serve the public less well in the ninth term as in the 1st term; even if quality of service remains constant, incumbency advantage implies that it is difficult to replace an incumbent with a superior challenger.
A central problem with all of these arguments is that the people who are being implored to vote for term limits are the same people who reelect their representatives. Ultimately, it is the voters who are responsible for incumbency advantage. Term limit proponents treat incumbency advantage as if it were an exogenous force imposed upon the voters by a nefarious power. They often fail to ask, "If incumbency advantage creates unresponsive politicians why do voters continue to reelect their representatives?" And if the question is asked at all, the answers given are unconvincing. Most arguments for term limits gloss over this issue and for the moment I shall follow this practice. After surveying a broad spectrum of arguments for term limits I shall return to this problem and analyze it in detail. Surprisingly, a proper understanding of the problem leads to a new conceptualization of term limits and their benefits.
With these preliminaries on incumbency advantage kept in mind let me turn to the arguments for term limits.
A New Political Man?
Edward Crane (1990) and James Coyne and John Fund (1992) argue that term limits would remove the professional politician from Washington and bring back the superior "citizen-statesman" and the "citizen legislature." They argue that term limits would make politicians more representative and more responsive.
The main argument that suggests term limits will make politicians more representative is that a term-limited political system will attract or "select" different types of politicians than an unlimited system. In an unlimited system most politicians who run for Congress are looking for a career. In a term limit system most politicians who run for Congress will be looking for, at most, a 12-year job. [2] Career politicians, it is assumed, have different preferences than non-career or citizen-politicians. This statement is rarely justified but the reasoning seems to be that a non-career politician will be more aware of the results of his actions on the private sector. Coyne and Fund (1992: 17) approvingly quote George McGovern, formerly the 1972 Democratic presidential candidate and now a hotel owner, who says, "I wish I had known a little more about the problems of the private sector . . . I have to pay taxes, meet a payroll--I wish I had a better sense of what it took to do that when I was in Washington." Reduced to essentials, the selection effect argument says that term limits will create a more diverse legislature and, in particular, a legislature with more business experience.
Historically the House was not more diverse when turnover was high. During the 19th century, when turnover rates were four to six times greater than they are today, lawyers made up over 60 percent of the House, more than at any other time before or since. The occupational background of representatives has remained remarkably stable since the founding of the United States. Lawyers have always made up about 50 to 60 percent of the House and since the 1850s those with business backgrounds, the largest group after lawyers, have constituted about 20 percent of the House (Bogue et al. 1976). Observed changes in turnover are much greater than those which would be created by term limits. No argument has been made which explains why changes created by term limits should lead to greater effects on the diversity of the House than have been observed from natural tenure changes in the past.
Most characteristics of representatives have been as stable across historical tenure changes as has been their occupational background. Those characteristics which have changed have done so in pace with changes in the suffrage. If a citizen-legislator is a representative who shares similar characteristics with her constituents then it is clear that we have more citizen-legislators today than ever before. Not only are women and minorities better represented in the late 20th century but representatives are more likely to come from the same "class" as voters (as measured, for example, by education levels; Bogue et al. 1976). The political class of the 19th century had fewer citizen-legislators than today and the process as a whole was more restrictive. Perhaps these factors made the politics of the 19th century preferable to those of today but few people, term limit proponents or otherwise, appear willing to make this argument.
James Payne (1991b) argues that term limits will make politicians more responsive because long-tenured politicians become "socialized" to big spending. He notes that at congressional hearings witnesses advocating government spending overwhelm their opponents by a ratio of 145 to 1. Special interests provide lots of information about the benefits of spending but little information about the costs. Similarly, the pressure to go along with spending increases is intense, the pressure to cut spending is diffuse. Faced with this imbalance of information and pressure Payne argues that representatives become "socialized" to higher spending and special interest influence. [3] In support of his argument, Payne (1991a) presents evidence that spending increases with seniority. That is, regardless of party affiliation, a senior representative tends to vote for more spending than a junior colleague. [4]
The socialization argument is not the only explanation for a positive relationship between spending and seniority. Political power, for example, increases with seniority, so higher spending could be due to more power rather than more seniority per se. If this explanation of the spending-tenure correlation is correct then term limits would sunder the relationship between seniority and power but would leave unaltered the more fundamental relationship between power and spending. With more politicians coming to power sooner, term limits could increase spending.
Robert Reed and Eric Schansberg (1993) carefully distinguish between the different arguments for a spending-tenure relationship and they look for indications in the data which would suggest one explanation is better than the other. Unfortunately, when the data is split into Republican/Democrat and House/Senate components they find that no single hypothesis can explain the data. One hypothesis works better for Democrats, another for Republicans and so forth. The findings of Reed and Schansberg (1993) call into question the existence of a true relationship between spending and tenure.
Neither of the arguments for term limits and the new political man are persuasive. The occupational composition of the legislature is unlikely to be greatly affected by term limits. More work on the spending-tenure hypothesis would be useful. However, with the evidence "to date," this hypothesis is shaky ground on which to base support for term limits.
Interest Groups and Term Limits
Proponents of term limits argue that term limits will reduce the power of special interests. The citizen-legislator hypothesis for this effect has already been discussed. A more plausible argument is that term limits increase the cost of bargaining and making commitments. As these costs increase the amount of special interest legislation is reduced.
Since political exchange is rarely simultaneous or subject to enforceable contracts there is potential for opportunism. Will the representative exert high effort for the interest? Will the interest fulfill its promises of support? Answering these questions in the affirmative requires that there be mutual trust between interest and representative, which can only be created in a long-term relationship. [5] Similarly, log-rolling requires that representatives know and trust one another. Term limits will increase the turnover rate for committee chairs and indeed for the entire legislature--every 12 years the legislature will be composed of entirely new representatives--thereby making trust difficult to build up. Long tenure also allows committee chairs to develop "property rights" in certain areas of legislation (Weingast and Marshall 1988). With term limits, property rights in policy areas will be more difficult to establish and this will make it difficult for politicians to logroll and trade political support.
Furthermore, even if term limits are long--potentially long enough for trust to be created--the fact that the final period is known can disrupt the evolution of cooperation. In a prisoner's dilemma type game if the players know the final period the cooperative strategy can never be rational. In the final period of the game defection is rational and backwards induction show that this is also true of the first period (Kreps 1990). Many problems that representatives face are like a prisoner's dilemma: it is better for each representative to defect from a political bargain rather than cooperate but when both representatives defect both are worse off. Factors other than an uncertain final period can make cooperation rational so this result does not imply that rational cooperation will disappear with term limits. [6] Nevertheless, it is easy to see that representatives in their final terms will be more tempted to break bargains than those in their first term and that this will make cooperation more difficult.
In addition to making it more difficult for representatives to bargain with interest groups, as well as among themselves, term limits increase the costs of committing the legislature through time. Legislation passed in the 98th Congress is unlikely to be overturned by the 99th Congress if the large majority of representatives are reelected. Legislation is more likely to be overturned when turnover rates are high. Although the possibility of overturning reduces the total demand for special interest legislation the number of bills passed may actually increase. [7]
These factors all work towards reducing the power of special interests. Lengthy terms, however, are only one way of producing long-term bargains. Substitutes exist, and these will be used increasingly in a term limit world. Political parties, for example, can substitute for lengthy terms in the production of long-term bargains. Instead of working through individual representatives special interests could lobby through the parties. Political parties in turn could use their growing power and wealth to discipline politicians. Politicians today have independent power bases and can afford to ignore their parties. Without lengthy terms, independence becomes harder to acquire and the resources of a stable, long-lived, political entity like a political party become more valuable and less easy to ignore.
The same forces which increase the cost of special interest legislation also increase the cost of producing public goods. Public good production requires trust and long-term bargains. A reduction in the power of special interest groups because of an increase in the cost of making bargains is not, therefore, necessarily welfare enhancing. Moreover, since public goods are probably already underprovided (the free rider problem exists in politics just as it does in markets), a reduction in public goods production is likely to be more costly than the benefit from reduced special interest legislation. In a model developed by Tyler Cowen, Amihai Glazer, and Henry McMillan (1994) private (special interest) goods must be tied to public goods in order to provide an incentive for politicians to produce public goods. The public good of national defense, for example, is provided because military contractors have an interest in lobbying for national defense. If this model is correct, a reduction in special interest legislation is not necessarily to be applauded.
Finally note that greater rotation in office increases the incentives of politicians to pass legislation which benefits their own post-Congress careers. This is just the negative spin on the pro term-limit argument that term-limited politicians will be more concerned about the effects of their actions on the private sector.
In summary, term limits increase the cost of lobbying through individual politicians. Special interests have more than one means of lobbying, however, and in a post term limit world they will turn to lobbying through the parties and other more long-lived actors. The final effect on special interest politics is likely to be small. To the extent that special interest politics is reduced we should also be concerned that public goods production may be reduced.
|
|
? It's very likely he'll get a second term considering his approval rating ... 
