"An Event History Analysis of State Legislative Term Limits"
By Dr. Bruce Bordner
University of Virginia
Abstract
This study utilizes descriptive event history analysis to examine the adoption of state term limitations in the United States. The study examines under what conditions states have adopted term limitations for state lawmakers. The study examines: structural factors in the form of the presence or absence of the initiative process; internal political factors in the form of partisanship within state governments; legislative turnover; external political factors in the form of voting trends; economic conditions; and, a geographic "border state effect." Support for structural factors, external political factors, and the geographic "border state effect" were found to be associated with the conditions under which states adopted term limitations. Internal political factors and legislative turnover both were found to have displayed little association with term-limitation measure adoption. Furthermore, the impact of overall state economic conditions was not significant, as voters do not appear to hold individual state lawmakers and their state legislature accountable for the state of their state's economy, but this may be due to voter inability to assign responsibility to one political actor or party.
Introduction
At present, some nineteen states possess term limitations restricting the lengths of service of state legislators. Two additional states, Massachusetts and Washington, passed term limitation measures only to see these measures struck down by the courts. While significant research has been completed on the subject of term-limitations, little research appears to have conducted on the conditions under which states have adopted term limitations on state lawmakers.
The research in the area of term limitations on the state level is slowly accumulating as more states adopt term limitations on state lawmakers (Benjamin and Malbin 1992; Donovan and Snipp, 1994). At present, little seems to be known about the reasons individuals support term limitations for lawmakers (Will 1992; Donovan and Snipp, 1994). Many of the analyses and studies completed on term limitations address the constitutionality of the limitations and the difference between state and federal limitations (Fett and Ponder 1993). Recent court rulings provide additional information in this area.
Partisanship can also play a role in how voters make their decisions on term limitations. If a voter perceives that a term limitation measure may weaken the control of an opposition party in the legislature, support of the measure may be more desirable. In the adoption of term limitation in California in 1990, reapportionment may have combined with partisanship. Democrats had controlled the state legislature and Republicans had controlled the governorship for many years. The strength of the Democrats in the legislature had been attributed to the enactment of a favorable reapportionment in 1980 (Donovan and Snipp, 1994). Lastly, studies have found that voters hold state governors and their party responsible for economic conditions (Niemi and Stanley, 1995), but not necessarily individual legislators.
Other research, however finds that it is not voter dissatisfaction with state legislators, specific incumbents, or political party ideology, but rather voter cynicism and self interest that drive the adoption of term limitations (Karp 1995). But, this voter cynicism must be combined with the ability of voters to act to enact term limitations on state lawmakers. Therefore, the presence or absence of the state initiative process may prove significant to these efforts.
The impact and results of term limitations on legislative turn-over have also received significant study (Kurtz 1992; Price 1992; Copeland 1992; Everson 1992; Moncrief et al., 1992; Opheim 1994; Caress 1994; Reed and Schansberg 1994; Chadha and Bernstein, 1996). Lastly, Chadha and Bernstein have examined the rationale behind different limitations being adopted in different states (Chadha and Bernstein, 1996).
Term limitations are often proposed by small groups of political activists (Chadha and Bernstein, 1996; Rothenberg 1992). Proposals created by these groups of activists are put before to the public through the initiative process. The initiative process allows for one of two possible outcomes; the initiative is either approved or disapproved by voters. The public is not allowed to alter the language of the initiative (Chadha and Bernstein, 1996), but rather the public must accept or reject the measure, as is. The groups that propose the initiatives often believe that they stand to gain in terms of representation through the adoption of the term limitation initiatives. Research indicates that the "out party" is the group most likely to gain through the enactment of term limitations (Thompson and Moncrief, 1993; Jewell 1994; Clucas 1994; Chadha and Bernstein, 1996). Further research indicates that the impact of term limitations may be to restructure group relationships within state legislatures (Moncrief, et. al., 1992). Guerra has even argued that under-represented minorities might gain in representation due to legislative turn-over caused by term-limitations (Guerra 1991). Lastly, studies have found that support for term limitations is stronger among women and younger voters (Donovan and Snipp 1994).
Statewide initiatives often require the signatures of many thousands of voters in order to qualify for state ballots and are subject to strict deadlines. In some states, multiple initiatives on a single subject may appear on the ballot at the same time. As such, it is possible that voters may be required to make decisions from among competing measures that address the same topic from a variety of angles (Banducci 1992). Furthermore, the amount of information available to voters at the time decisions are made to support or oppose a measure can impact the voter's decision making process (Magleby 1984; Zisk 1987).
This study has chosen to examine the adoption by states of term limitations on state legislators for the reason that all state adopted limitations, to date, on federal legislators have been voided by the federal courts. The history of term limitations is not short. The origin of term limitations in the politics of the United States dates back to the Articles of Confederation in 1781. Under the Articles of Confederation, representatives to the national legislature were restricted to serving three years in six. Governance under the Articles of Confederation did not last long. The Articles of Confederation did not provide for a strong national government and were deemed by many individuals at the time to be greatly flawed, thus leading to a new Constitution in 1789. Unlike the Articles of Confederation, the new United States Constitution did not place limitations on Members of the United States House of Representatives, United States Senators, or on the President. Term limitations at the federal level did not reappear until the Twenty-Second Amendment was enacted in 1951. This amendment prohibited an individual from being elected President of the United States more than twice. The amendment also allowed an individual assuming the Office of President the ability to finish the term during which they succeeded to the office and to seek re-election so long as they did not serve more than ten years as President. The next discussion of federal term limitations occurred in 1978 before a United States Senate Judiciary Committee sub-committee. In the end, this sub-committee did not take any action on the issue. The topic resurfaced in 1988 when the Republican Party platform included a plank favoring term limitations for the United States Senate and the United States House of Representatives.
In May of 1995, the United States Supreme Court ruled on the case of U.S. Term Limits v. Thornton, (115 S.Ct. 1942, 1995). Through this case, the Court ruled that states could not enact restrictions on federal office holders. States could not change the requirements, established in the United States Constitution, for the holding of federal elective offices. The result was the voiding of all state enacted restrictions limiting the terms of Members of Congress and the United States Senate. The Court's ruling in U.S. Term Limits v. Thornton did not void state enacted term limitations on state lawmakers.
Since 1990, twenty-one states have enacted term limitations for state lawmakers. Limitations in two of these states, Washington and Massachusetts, were thrown out by the courts. The surviving limitations vary in terms of both their length and content. In California, for example, lawmakers are limited to serving three terms in the California State Assembly and two terms in the California State Senate. The Governor is also limited to two terms. California's term limitations are lifetime limits. Other states, such as Massachusetts and Wyoming, limit lawmakers to a maximum number of years of service within at certain time span. For example, in Wyoming, state senators are limited to serving a maximum of three terms within a span of twenty-four years. Wyoming State Representatives are limited to six terms within a twenty-four year period. In a number of states, such as California, Arkansas, Arizona, and Florida, the laws limit elected members of the executive branch to two consecutive four-year terms. Maine limits members of the executive branch and other constitutional officers to a maximum of four consecutive two-year terms. A detailed accounting of the limits individual states have placed on state legislators can be seen in Appendix 'A' Table One.
What factors guide the adoption of term limitations on state lawmakers within the United States? Is the adoption of state term limitations influenced or guided by internal or external factors, or both? Are there structural variables present in states adopting limitations on state lawmakers that are not present in states not adopting term limitations? How prevalent are issues such as legislator turn-over, voter turnout, and the economy, in the adoption of state term limitations?
What this article present is a descriptive event history analysis on the adoption of state term limitations. By descriptive, I mean that it attempts to understand under what conditions states have adopted term limitations for state lawmakers. Therefore, the construct of interest is the adoption of term limitations on state lawmakers within the United States. The independent variables in this study take five forms: structural, politically internal, politically external, economic, and geographic. The data utilized in this analysis has been brought together from a wide variety of reputable sources including, but not limited to, the National Conference of State Legislatures, the National Council on State Governments, and the United States Department of Commerce. In some cases, data for all twenty-one states were not available, thus limiting the analysis.
The Model
State Term Limitations: An Empirical Examination
Dependent Variable:
For purposes of this study the observed enactment of term limitations by a state will serve as the dependent variable. Enactment is signified by a dummy variable with the numeral "one" denoting a state's enacting term limitations and the numeral "zero" signifying the absence of enactment of the policy. The break down of this data is contained in Appendix 'A' Table Two. A total of seven hypotheses will be set forth and tested within this event history analysis. Each of these hypotheses will represent a different independent variable.
Unit of Analysis:
Because this study seeks to examine the factors involved in the adoption of state legislative term limitations in the United States, the forty-eight contiguous states of the United States will constitute the unit of analysis. The reason for selecting this unit as the basis of our analysis is clear. The two states excluded, Alaska and Hawaii, do not border any other states of the United States. Data for these two states will, however, be included in the discussion of average state voter turnout.
The dependent variable in this examination is the adoption of term limitations on state lawmakers. The study makes no distinction between states that adopt term limitations on multiple occasions and states in which state adopted term limitations have been voided by state courts. The issue this study is concerned with is the act of adopting limitations on the terms of state lawmakers. For example, limitations enacted by state voters in Nebraska are counted in the total number of states enacting term limitations on state lawmakers, even though the Nebraska State Supreme Court has declared these limitations void, only to see term limitations re-enacted by voters in a different form.
The time period covered in this study begins with the founding of the United States of America and continues to the present. This has permitted the author to gather data on the adoption by states of term limitations on state lawmakers for the entire history of our nation, thus the data collection is fully comprehensive in this regard.
Independent Variables
Structural Variables:
Structure of State Constitutions: The power of the people to adopt legislation through an initiative process amounts to a granting to the people of the power to legislate. Who legislates determines what policies are adopted. In Southern Politics, V.O. Key, states that the real problem in southern politics is, "the maintenance of control by the white minority." For over two-hundred years, race relations have guided the politics of the south. The dominance of race as a political issue in the south can be seen in the scars left by its history of discrimination. The early American south was dominated by a slave based agrarian economy. Blacks were not made citizens or given the right to vote until after the Civil War, but even then they faced polling taxes, literacy tests, and other efforts designed to deny them the franchise. All of these efforts go to the heart of the maintenance of power in the south by V.O. Key's white minority. Many southern states do not give citizens the ability to adopt legislation through the initiative process. It seems logical to conclude that the reason for this is the desire to control power within the political system. In a number of western states, however, more progressive state constitutions have granted the people the power to enact laws through an initiative process. Race and class do not seem to have been strong factors in these states.
Voice of the People. This independent variable seeks to examine the manner in which the people act in place of their government's by enacting term limitations on state lawmakers. Also, it seeks to determine when populist sentiments are exhibited when government fails to act for one reason or another.
Politically Internal Variables:
Divided Government. This variable examines the level of political party competition within individual state governments. The partisan control of the two houses of the state legislatures and of the governorships of the states are examined in order to study these institutional conditions. Nebraska is excluded from the analysis of this variable because it possesses a non-partisan, unicameral, state legislature.
Three major theories have been offered to explain divided government on the federal level. James Sundquist has advanced the belief that people should want an end to divided government because it would lead to quick legislative action by the government and executive. David Mayhew has advanced the view that divided government really does not matter; that policy will or will not be enacted no matter who is in charge of the government. A third belief is that people prefer divided government as political check within the federal government. Lastly, Charles O. Jones argues, in The Presidency in a Separated System, that perhaps it is not possible to have a unified national government in the United States because both the Congress and the President represent different and competing interests. Members of Congress represent the interests of the constituents in their districts, whereas the President represents the interests of the entire nation. It is possible to transfer many of these arguments to the state level in order to explain the occurrence of divided government. Adapting the Sundquist argument to the state level, people should want unified state government because that will lead to an efficient legislature. According to the Mayhew approach, it should not matter whether the partisanship of the state government is divided or not because legislation will still be enacted. Additionally, adapting the third view holds that people may prefer that one house of a state legislature check the other house of the legislature or that both houses of the legislature be controlled by members of one party and the governor be of the opposite party. Lastly, moving the Jones argument to the state level, state lawmakers represent similar competing interests that may not be the same as those of the state's governor, thus leading to divided government. Which view is correct seems open to debate on the state level as much as it has been debated about the federal level.
The issue of divided government is utilized as a variable in this analysis because it indicates the political partisan unity within state governments. Simply put, the author seeks to determine if divided government is a factor common to the adoption by states of term limitations for state lawmakers.
Turnover. This variable seeks to examine if anger or a lack of confidence exists among the people of states toward their state government. Turnover indicates what percentage of a state's legislature that is not returned to office in a given year. If turnover is low, but voters seek to enact term limitations, perhaps the people of that state seek to counter the powers of incumbent politicians to be re-elected by forcing the election of new lawmakers.
In "Institutions, the Economy, and the Dynamics of State Elections," John Chubb wrote about the power of incumbency and the institutionalization of state legislatures. He states that, "[i]t is known (see, e.g., Calvert 1979; Ray 1974, 1976; Rosenthal 1974; Shin and Jackson 1979) that member turnover in state legislatures declined, albeit unevenly, during the period 1940-80, as more incumbents chose to make the legislature a career." Chubb continues, "[b]ut this is only a part, and in fact an early part, of the process of institutionalization. Institutions automatically acquire stability when members decide that they value and want to remain in them. But that stability is a long way from that which institutions acquire when members find ways to protect their positions against outsiders - that is, electoral challengers- who also find those positions attractive and can bid for the support of voters. Legislatures do not become thoroughly institutionalized, therefore, when legislators decide they want to keep their positions but[,] when they demonstrate that they can." An examination of turnover should help to determine if it is a factor present when term limitations for state lawmakers are adopted. The adoption of term limitations might serve as a means for overcoming this institutionalization of legislators in their positions.
Politically External Variables:
Turnout. This study draws upon voter turn-out as an indicator of the degree of political participation by individuals within states. Data from all fifty states is utilized to identify if states adopting term limitations for state lawmakers posses higher degrees of political mobilization than the average of all states. This study's use of voter turn-out is similar to Paul Peterson and Mark Rom's use of voter turn-out data in "American Federalism, Welfare Policy, and Residential Choices," as an indicator of socio-economic class. Their belief was that "States with high voter turnouts probably have higher class mobilization..." The argument in the present study is that higher class mobilization, demonstrated in higher voter turn-out in non-presidential general elections, serves as an indicator of the increased desire on the part of voters to participate in the political system, and to have this system responsive to their needs and interests. This is in keeping with Peterson and Rom's view that one party or another would be responsive to the concerns of voters for broadening welfare.
Economic. The role of economics in the election of lawmakers and the holding of lawmakers accountable through elections has been examined in numerous studies, but many questions seem to remain. Do voters base their votes based upon economic conditions of the nation or of their state or both? The author takes an approach similar to that taken by Paul Peterson and Mark Rom in "American Federalism, Welfare Policy, and Residential Choices." In that study, the authors utilize an examination of economic conditions as indicators of the ability of lawmakers to "react slowly to environmental changes and the economic factors would influence policy gradually over a five year period." The present study utilizes data covering three years for the same purpose. The study seeks to determine if economic factors serve as a motivation for the public to assign blame to lawmakers in the form of term limitations.
Geographic. The role of geographic proximity in the adoption of policies by governments has been covered by a number of political scientists. In "State Political Culture," Erickson, McIver, and Wright argue that the political culture or attitudes that a state possesses may influence political attitudes. What they found was that demographics were not enough to determine the variation between states. While demographics do not explain the political culture, they did find some relationship between regional factors. They found that, "southern and border states show party effects that are considerably more Democratic than the norm. Southern and interior western states show ideology effects that are more conservative than the average." They conclude that only a small portion of the ideological variance is accountable as a regional effect. They also examine "Social Context" in order to study the, "...hypothesis that peoples' attitudes are influenced by the aggregated attitudes of those around them." What they find is that the, "...social context might be examined, for example, by seeing whether neighborhood effects on the vote correlate with the vote predicted from residents' individual characteristics (Huckfelds 1979, 1984). At the state level, the social-context hypothesis offers an obvious prediction. If state effects reinforce state attitudes that result from individual-citizen characteristics, then the demographic component of state opinion (due to individual characteristics) should be related to the cultural component (due to residence in the state apart from individual effects.)" While they do not find that cultural or demographic variables help to explain state partisanship or political ideology, they do find a role for the state effect. They find that, "... the geographic sources of the variation in political opinion measured in this paper may extend beyond discretely different cultures separated by state boundaries. 'State' effects captured by state boundaries may represent in part the sum of the effects of regional variation within states or even of variation that transcends state lines." They found an argument for movement or variation across state lines - this would presumably include the movement of policy ideas.
In "State Lottery Adoptions as Policy Innovations: An Event History Analysis," Frances Stokes Berry and William D. Berry, make the argument that a state is more likely to adopt a policy innovation if a nearby leader state has adopted that policy. They state that, "[t]his definition is consistent with research that has found that there are states to which other states in a region look most frequently for innovation ideas (Grupp and Richards, 1975; Menzel and Feller 1977)." They continue, "[t]his conception of regional diffusion is most attractive when there are reliable data about which states are perceived by public officials to be regional leaders in a policy area."
The structure of the border state definition utilized in the present study is the same as was employed by Berry and Berry in their event history analysis. It is also consistent with Erickson, McIver, and Wright's view of variation transcending state boundaries. My argument is that when the voters of one state in a region adopt term limitations, other states will soon follow if voters desire the same power or control over state lawmakers and can act upon this desire. Furthermore, when one state has a strong political movement that allows it to pass a statewide initiative limiting the terms of lawmakers, it then becomes easier for people to bandwagon on to this successful idea in neighboring states that possess the initiative process.
Hypotheses
The first two hypotheses are structural in nature.
Hypothesis One:
The first hypothesis seeks to examine a structural issue relevant to the adoption of term limitations on state lawmakers. Term limitations for state lawmakers can be enacted into law in two forms: (1) legislation passed by a state legislature and signed into law by that state's governor, or enacted over a governor's veto; or (2) enactment by that state's voters in the form of a state initiative. This study will examine the twenty-one states adopting term limitations on state lawmakers in order to determine the method of adoption that is utilized most often. It hypothesizes that the enactment of state term limitations on lawmakers will be more likely to occur through the second form, a state initiative, than through the first form, legislative action. The basis for this hypothesis is that while voters may desire to limit the terms of lawmakers, lawmakers are unlikely to act to limit their own terms.
Hypothesis Two:
The second hypothesis, perhaps a corollary to the first, seeks to examine the occurrence of the adoption of term limitations on state lawmakers by looking at those states that possess the initiative process and those in which such a process is absent. It is hypothesized that the adoption of term limitations should be significantly diminished in those states not possessing the initiative process and present at a much higher rate in those states possessing the initiative process. The basis of this hypothesis is that if a state does not possess the initiative process, then voters, even if they desire to enact term limitations on state lawmakers, are unable to enact term limitations.
The third and fourth hypotheses are politically internal in nature.
Hypothesis Three:
The third hypothesis examines the partisan nature of the state governments of all states enacting term limitations. It is hypothesized that term limitations on state lawmakers are more likely to be enacted under divided state government, than under a united government. The argument in favor of this hypothesis rests on the fact that government is more likely to function properly when a government possesses partisan unity, than if control of the government is divided. Additionally, voters will enact term limitations when they seek to be able to assign responsibility within government to one group of people or another. Divided government makes it more difficult for voters to assign this responsibility, thus the greater need to enact term limitations under divided state governments, than under unified state governments.
Hypothesis Four:
The fourth hypothesis examines the turnover rates of lawmakers. It is hypothesized that turnover rates will generally be low in the forty-eight contiguous states. The absence of turnover in a state legislature combined with the presence of the initiative process provides voters with a desire and opportunity to enact term limitations. The reasoning behind this hypothesis is that voters seek to overcome the powers associated with incumbency by enacting term limitations.
The fifth hypothesis is politically external in nature.
Hypothesis Five:
This hypothesis examines voter turnout to see if voter turnout is higher in states enacting state term limitations than turnout among all fifty states. The basis of this hypothesis is that higher turnout increases the likelihood that term limitations will be adopted by turning out not just the core base of voters, but more disenchanted voters, as well. In general, states enacting term limitations for state lawmakers should be more politically motivated. A higher rate of voting in non-presidential general elections will serve as the measure of the presence or absence of this factor.
The sixth hypothesis is economic in nature.
Hypothesis Six:
The sixth hypothesis examines what role, if any, is played by state economies in the adoption of state term limitations. It hypothesizes that, similar to previous studies, state term limitation adoption rates will not be higher in states with low growth rates in gross state product over the immediate three year period prior to adoption than in states with higher rates of growth in gross state product. The basis of this hypothesis is that voters do not generally hold individual state lawmakers accountable for the state of their state's economy. Rather, state voters tend to assign overall responsibility for the economy to a governor, or, even, the President.
The seventh hypothesis is geographic in nature.
Hypothesis Seven:
This hypothesis seeks to examine what regional influence is present in the adoption of term limitations on state lawmakers. It is hypothesized that the adoption of state term limitations will be greater when bordering states have previously adopted term limitations or adopt them in the same election. The basis of this hypothesis is that the previous adoption of term limitations on state lawmakers and/or the simultaneous enactment of such measures have a rallying affect on neighboring states.
Sources of Data
Data for this study has been gathered from a variety of sources. Data for the first four hypotheses has been gathered from the National Conference of State Legislatures (NCSL). NCSL provided data breaking down states possessing the initiative process versus states without the initiative process, data noting the breakdown of the method of adoption of term limitations in those states having adopted limitations, data on the partisan breakdown of state governments, and data on legislator turnover within states. The voter turnout data for hypothesis five has been gathered from the Book of the States, 1994-95 and from the United States Department of Commerce in the Statistical Abstract of the United States, 1995. Economic data for gross state product in hypothesis six was provided by the United States Department of Commerce in the Statistical Abstract of the United States, 1996. Lastly, the author took the NCSL list of all states adopting term limitations for state lawmakers and compared it with the list of border states, as defined in Appendix 'B' Table One, to examine the border state effect of hypothesis seven.
Results
The model I have examined provides a unique portrait of the adoption of state legislature term limitations within the United States. An examination of the adoption patterns indicates that three states enacted term limitations on state legislators in 1990. In 1992, another ten states enacted similar limitations on state lawmakers. Maine was the only state to enact term limitations on state lawmakers in 1993. In 1994, term limitations were enacted by another six states. Louisiana's state legislature was the only state enacting term limitations on state lawmakers in 1995. This was done when voters approved a legislative proposed state constitutional amendment in the form of an initiative. This break down can be seen in Appendix 'A' Table Three.
The data provide strong support for the first hypothesis. Table Three indicates that a total of twenty of the twenty-one states that have enacted term limitations for state lawmakers enacted them through a state initiative process. Utah is the only state that has enacted term limitations through the traditional legislative process. Term limitations did appear on the Utah state ballot later that same year, but were defeated as the issue of term limits had already been resolved by the Utah State Legislature's prior action. The twenty of twenty-one states adopting term limitations through the initiative process provides strong support for the hypothesis that term limitations are more likely to be enacted by a state's voters than by a state's legislature.
The data also provide strong support for the second hypothesis. The adoption of term limitations for state lawmakers was higher for states possessing the initiative process and virtually absent in states not possessing the initiative process. Appendix 'A' Table Four indicates that a total of twenty-four states possess the initiative process. A comparison between the data in Appendix 'A' Tables Three and Four indicates that twenty of twenty-four states possessing the initiative process, or slightly more than eighty-three percent, had enacted term limitations on state lawmakers by November 1996. Of the remaining twenty-six states in which the initiative process is absent, only one state, or slightly under four percent, has adopted state term limitations. This is strong evidence that the adoption of term limitations for state legislators is highly related to the ability of voters to enact legislation and state constitutional changes through an initiative process. In states lacking the initiative process, legislators have little to fear from term limitations because they are unlikely to limit their own ability to be re-elected.
The data show weak support for the third hypothesis. Of the twenty states adopting state term limitations within partisan political systems, a total of eleven or fifty-five percent, possessed divided government, while nine or forty-five percent possessed unified government. Again, Nebraska was excluded from this calculation because it possesses a non-partisan state legislature. The partisan breakdown of state governments for states enacting term limitations on state lawmakers can be seen in Appendix 'A' Table Three. A further analysis of the data indicate that the likelihood of adopting term limits was slightly greater when either control of the state legislature was divided among more than one political party or the state legislature possessed unified control by a political party other than that party controlling that state's governorship. In those instances in which term limitations were enacted with a united governor and legislature, Democrats dominated state houses and governorships by a two-to-one margin. Additionally, when political leadership of the legislature and governorship was divided among different political parties, nearly eighty-two percent possessed Republican governors and divided legislatures; whereas, only slightly more than eighteen percent of those governments possessing divided governments possessed Democratic governors and Republican legislatures. The data also indicate that term limitations were more likely (76%) to be enacted when state legislatures were controlled by a single political party than when control was divided among one or more political parties. The conclusion can also be made that term limitations are more likely to be adopted with Democratic governors and unified control of state legislatures by either political party or when Republicans dominate the governorships and Democrats dominate the state legislatures.
In 1990, the first three of the twenty-one states adopted their term limitations on state lawmakers. A total of one-hundred percent of those states adopting term limitations in 1990 possessed divided legislatures and Republican governors.
In 1992, fifty percent of the states adopting term limits possessed state legislatures in which political power was divided. A total of forty percent of states adopting term limitations that year possessed divided political power in their state legislatures and had Republican governors, whereas only ten percent had Democratic governors. Slightly more, thirty percent, of state governments possessing unified political leadership in their state legislatures and governor's houses were dominated by Democrats. A total of sixty percent of those states adopting term limitations in 1992 had Republican governors.
A single state, Maine, enacted term limitations in 1993. Maine possessed a divided government led by a Republican governor.
1994 saw five states enact term limitations. A total of forty percent of those states enacting term limitations on state lawmakers possessed divided governments headed by Republican governors. No Democrats were governors of states possessing divided government and enacting term limitations. Where state governments were found to be dominated by a single political party, twenty percent of those states enacting term limitations were led by Democrats, while forty percent were led by Republicans. A total of eighty percent of states enacting term limitations in 1994 had Republican governors.
The data indicate little support for the fourth hypothesis that turnover in state legislatures will be generally lower in states adopting term limitations for state lawmakers. The data in Appendix 'A' Tables Five and Six indicate that the average turn-over rates among all fifty states is fairly low. Average state legislative turnover for all states from 1986 to 1990, shown in Table Six, was 22.65% versus 21.45% for states possessing term limitations, shown in Table Five. In fact, the turnover rate for all states was slightly higher in 1986 (22.48%) than for states that would later enact term limitations (22.62%). The average state turnover rates for states that have enacted term limitations ranged for the time period 1986 to 1990 from a low of 10.33% in Ohio and California at 13% to 39.33% for Oregon. Furthermore, an examination of turnover data from 1979 to 1989 indicates that the general differences among state legislature turn-over may be the result of the degree of professionalization and institutionalization of each state legislature. Data from the National Conference of State Legislatures indicate greater turn-over in both state senates and state assemblies that are part-time, have smaller staffs, and lower pay. Whereas, the more professionalized state legislatures tend to possess larger staffs, be full time, have higher pay, and tend to have less turn-over. Specifically, their data show that sixty percent of state senates and sixty-five percent of state assemblies in the full time, professional legislature, turned-over from 1979 to 1989 or approximately six percent of the state senates and six and a half percent of the state assemblies per year. Meanwhile, fully eighty percent of non-professional state senators and state assemblies turned over from 1979 to 1989 or about eight percent per year. The conclusion can be made that a low level of state legislative turn-over is present in all fifty states, but turn-over is greater in states possessing less professionalized state legislatures. This low level of turn-over only becomes a factor in the enactment of term limitation for state lawmakers when the initiative process is present.
The data indicate significant support for the fifth hypothesis. The data indicate that the average state turnout for non-presidential general elections was significantly higher for those states enacting term limitations on state lawmakers than among non-enacting states. The data also indicate voter turnout rates in 1982, 1986, 1990, and 1994 of 64.80%, 60.12%, 60.53%, and 75.52% for state enacting term limitations versus 55.97%, 53.10%, 54.98%, and 70.05%, for non-term limitation enacting states. The conclusion that can be made is that there is a significantly higher degree of political motivation among voters in those states that eventually enacted term limitations than among the states not enacting term limitations. Lastly, the data indicate few differences among voter turnout within states from one election of the same type to the next. Additionally, in elections where voter turnout is generally higher or lower for a state, this increase or decrease in voter turnout appears similar across the entire data set. Turnout did not, however, show a marked increase in elections in which states adopted term limitations on state lawmakers.
The analysis indicates support for the lack of an economic basis for the adoption of state term limitations proffered in the sixth hypothesis. Economic data for gross state product was examined for those states enacting term limitations in 1992. The data for this analysis are contained in the Appendix 'A' Table Eight. Data collected are for the year in which term limitations were enacted and the two previous years. A comparison of the three year average for the ten states from 1990-1992 indicate growth rates ranging from .21% per year for Ohio to 5.64% for Oregon. Additionally, the average yearly growth rate for all ten states was 2.12%. Because the economic data range from fairly small, .21%, to a larger more sizable amount, 5.64%, with an average being an arguably reasonable, but not startling, rate of growth, 2.12%, then it seems plausible that the immediate economic conditions may not have been a major factor in the adoption of term limitations on state lawmakers. This result is consistent with the hypothesis that state voters will not hold state lawmakers accountable for the general state of that state's economy because voters are generally unable to assign responsibility for the economy to a single individual, such as a state's governor.
Lastly, the data indicate strong support for the seventh hypothesis. There does appear to be a regional influence among border-states in favor of enacting term limitations on state lawmakers. For the purposes of this study, only data for the forty-eight contiguous states were utilized. A table of the forty-eight contiguous states and the list of states bordering on each of them is contained in Appendix 'B' Table One. Data on states were examined for each of the five years in which the term limitations were enacted for state legislatures: 1990, 1992, 1993, 1994, and 1995. Appendix 'B' Table Two contains a break down of initiative possessing states and their border-states. The "border effect" will be deemed present if "State A" adopts term limitations on state lawmakers and it borders on another state that has either previously adopted term limitations on state lawmakers or does so in the same election year as "State A." By utilizing previous conditions providing for previous adoptions or concurrent adoptions, this study should be able to capture the presence of any transitive or bandwagon factor among the states.
In 1990, two of the three states bordered on each other with Colorado bordering on Oklahoma; therefore, a total of thirty-three percent of those states enacting term limitations bordered on another state enacting term limitations on the state level. A total of twenty-percent of states bordering on Colorado enacted term limitations in 1990, thirty-three percent of states bordering on Oklahoma, and no states bordering on California.
In 1992, the study found that three of the seven states that border Arkansas possessed the initiative process, and that two of them had previously enacted term limitations or enacted them that year.
The data indicate that in 1993, Maine was bordered on by two states, one of which was Massachusetts which had previously enacted term limitations (See Note on Table One of Appendix 'B').
In 1994, five of the six states possessing the initiative process and enacting term limitations had all of their border states enacting term limitations on state lawmakers in the 1994 elections or prior. In fact, slightly more that eighty-three percent of states possessing the initiative process and bordering on Utah were to enact term limitations on state lawmakers by the end of 1994 when Utah's legislature enacted the sole state legislature enacted measure limiting the terms of state lawmakers.
A careful examination of the data in Appendix 'B' Table Two also indicate that the likelihood of border states having an impact upon neighboring states with regard to the enactment of term limitations for state legislators is directly related to the ability to enact such limitations via the initiative process. Evidence of this fact can be seen in the one-hundred percent of Massachusetts' border states possessing the initiative process enacting term limitations, whereas a total of sixteen and two-thirds percent of the total states bordering on Massachusetts enacted term limitations. In fact, the number of bordering states adopting term limitations on state lawmakers but not possessing the initiative process was zero. Overall, the border state effect was more prevalent, slightly less than fifty-six percent, in border states possessing the initiative process, versus slightly more that thirty-nine percent in all states. Clearly, the border state effect exists, but this effect is greater in states possessing the initiative process, than in all states in general.
Conclusion
The data contained in this event history analysis provide support for the presence of a number of the factors examined in this study. The data clearly demonstrate strong support for structural factors and external political factors being associated with the conditions under which states adopted term limitations. Internal political factors and legislative turnover both were found to have displayed little association with term-limitation measure adoption. Furthermore, the impact of overall state economic conditions was not significant, as voters do not appear to hold individual state lawmakers and their state legislature accountable for the state of their state's economy, but this may be due to voter inability to assign responsibility to one political actor or party. Furthermore, there is also strong support for the transcendence of ideas across state borders through the geographic "border state effect,". Political movements and policy ideas do not stop at state lines.
Perhaps the strongest factor present was that of the internal structural of whether a state possessed the initiative. Term limitations were almost exclusively enacted by voters in states possessing the initiative process. Only one of the twenty-one states enacting term limitations on state lawmakers did not possess this internal structural variable. The reasoning seems fairly simple: voters must be empowered to act before they can act. Additionally, once passed, term limitations, referendums, and recall mechanisms have never been revoked by the people, only by the courts. (Schmidt, 1989). Clearly, the data on legislative turnover indicate that state legislative turnover is fairly low in many of the fifty states. Furthermore, surveys continually demonstrate that voters have only "some" or "very little confidence" in state lawmakers. Therefore, if voters have low confidence in their state lawmakers and if they possess the power of the initiative process, then they tend to be willing to limit the terms of state lawmakers. In all save one state, term limitations have been adopted through the initiative process. And, term limitations have been adopted in twenty-one of twenty-four states possessing the initiative process.
Perhaps this study has brought to the forefront another question that needs to be asked. What is unique or characteristic of states that have adopted the initiative process? Since the founding of our nation, states have placed before their citizens legislative measures to be approved. In fact, soon after the signing of the declaration of independence, the founders asked the states to draft state constitutions for the approval of the people of their respective states. Over time, states have placed more legislation before the people. When one examines a map of those states in which the people possess the powers initiatives, referenda, and recall, one notices that these powers are not restricted to only one portion of the nation. Therefore, the involvement of the people in the legislative process is not new or isolated to one or more regions of the country. (Cronin: 1989). It is only the initiative that is heavy concentrated in the western United States. And, this may have been a result of the political development of the nation. Many of the western states came into the Union just prior to the beginning of the Progressive movement. As a result, many of the western states may have included mechanisms for popular participation in the legislative process in their state constitution by granting the people the power of the initiative. In that sense, the initiative is a part of the political culture of these states. This might explain why the initiative process is more prominent in the western states, than in some of the earlier established eastern states.
Another example of this can be seen in V.O. Key's Southern Politics. Key presents the argument that the white minority was attempting to prevent blacks from participating in the political process. This was a part of the southern social and political culture. Its basis was in the slave based agrarian economy of the old south. Southern politics was, and to a degree still is, about race. When one looks at the states that possess the initiative process, very few are in the south (only Georgia and Florida). It seems reasonable to assume that one of the primary reasons for this lack of participation in the political process is because whites controlled the legislative elections and did not want black participation. Therefore, many southern states did not grant to the people the power of the initiative. There is no way around it, race is a part of the political culture and legacy of the south. This may serve to partially explain why the initiative process is not very prominent in the south.
The main conclusion to the question posed seems to be that increased voter turnout may be both an artifact resulting from the populist movement and the development of the political and social cultural fabric of individual states, such as those of the south or the west.
Another important result from this study stems from the higher rate of political involvement by voters in states possessing the initiative process, than in those states not possessing the initiative process. As was indicated in the discussion of voter turnout (see "Turnout" above), the increased rate of voter turnout seems indicative of an increased political motivation on the part of voters in initiative possessing states. Voters in initiative states had between a 5.55% and 8.83% higher turn out rates than voters in non-initiative possessing states for non-presidential general elections. The corollary is that voters in non-initiative states appear to possess a lower degree of political motivation. It seems reasonable that over time voters in states possessing the initiative process will become acclimated to the use of the initiative and its being involved in the policy making process. It is possible that voters in these states may have realized that they can impact the legislative process directly; whereas, the political culture in non-initiative states is not characterized by this direct political involvement. In essence, the voters may feel that their vote can make more of a difference in states that possess the initiative process because they can see that the measure on which they cast a vote either passed or failed. Voters in non-initiative possessing states lack this impact, thus they are unable to see the direct impact of their vote on the policy process. They only have an indirect impact on the implementation of policy initiatives through the election of lawmakers to represent them.
The border state effect is not limited by the initiative process. An example of this can be seen in the tax revolt that took place in California during 1978. An initiative, Proposition 13, was passed rolling back property taxes for California residents. The concept of a tax revolt spread to other states across the nation. The revolt was not limited to states possessing the initiative process. In states not possessing the initiative process, the policy idea was taken up by state legislatures. Therefore, policy ideas can originate as initiatives in one state and become legislative policy proposals in other states. Likewise, a policy proposal can originate in one state's legislature and spread to neighboring states where they may be pursued through either the normal legislative process or through the initiative process, where available.
The issue of term limitations differs from many policy initiatives. Term limitations threaten the power of lawmakers. Many policy initiatives do not pose such a direct threat to the livelihood of lawmakers. Therefore, term limitations will most likely not be enacted by state legislatures, even in neighboring states. This is consistent with the study's results. It is true that public pressure may be placed on lawmakers to support term limitations. It is also possible that candidates challenging incumbents may make support of term limitations a campaign issue. If enough candidates pledged to supporting term limitations are elected to a state legislature, then it becomes more likely that the state legislature might enact term limitations on itself. The caveat, however, is that while lawmakers may pledge to support term limits, they do not pledge to support a certain bill. Lawmakers sometimes support a policy idea only to find that the vehicle for implementing the idea is not one they can or choose to support. In other instances, lawmakers may pledge themselves to supporting term limits in order to satisfy a constituency, knowing fair well that the measure will not pass for a lack of votes. They may even change their minds when they get into office as to their position on term limits. All of this may serve to feed public cynicism, thereby fueling voter support for the enactment of term limitations on state lawmakers.
As astute students of political science, we realize that politics is all about power, with politicians seeking to control this power. When voters lack the power to legislate, the powers of institutionalization and institutionalized incumbency rule the day through politics as usual. Perhaps lawmakers in states possessing the initiative process should live by the adage, "Legislator Beware." In these states, the people rule.
References
Banducci, Susan, 1992. "Voter Confusion and Voter Rationality: The use of Counter Proposals in
the Direct Democracy Process." Presented at the American Political Science Association meeting, Chicago.
Berry, Francis S., and William D. Berry, 1990. "State Lottery Adoptions as Policy Innovations: An Event History Analysis," American Political Science Review,
Vol. 84, p. 395-415.
Benjamin, Gerald, and Michael Malbin, 1992. Limiting Legislative Terms. Washington, D.C. Congressional Quarterly Press.
Book of the States, 1994-5. Council on State Governments, Lexington, Kentucky.
Capell, Elizabeth, 1994. The impact of term limits: Early returns from California. APSA Legislative Studies Section Newsletter. Extension of Remarks 17(2):6-7,9.
Caress, Stanley, 1994. "The impact of term limits on state legislative behavior." Paper presented at the Southern Political Science Association Meeting in Atlanta, Georgia, November.
Ceaser, James. Professor, University of Virginia. A lecture from a course on the American
Presidency, spring 1997.
Chadha, Anita, and Robert A. Bernstein, 1996. Why Incumbents Are Treated So Harshly." American Politics Quarterly. Vol. 24, No. 3, July 1996, pp. 363-375.
Chubb, John (1988), "Institutions, the Economy, and the Dynamics of State Elections." American Political Science Review. 82:133-154.
Clucas, Richard, 1994. "The first post term limits election in California." APSA Legislative
Studies Section Newsletter: Extension of Remarks 17(2) 4-5, 15.
Copeland, Gary, 1992. "Term limitations and political careers in Oklahoma: In out, up,
or down. In Limiting Legislative Terms by Gerald Benjamin and Michael Malbin. Washington, D.C.: Congressional Quarterly Press.
Cronin, Thomas E., 1989. Direct Democracy: The Politics of Initiatives, Referendum, and Recall.
Cambridge: Harvard University Press.
Donovan, Todd, 1993. "The Social and Political Basis of Support for Legislative Term
Limitations." Presented at the annual meeting of the Western Political Science Association, Pasadena, California.
Donovan, Todd, and Joseph Snipp, 1994. "Support for legislative term limitations in California:
Group representation, partisanship, and campaign information." Journal of Politics,
56: 492-501.
Erickson, Robert S., John P. McIver, and Gerald C. Wright, (1987). "State Political Culture and
Public Opinion." American Political Science Review. 81: 797-813.
Everson, David, 1992. "The impact of term limitations on the states: Cutting the underbrush or
chopping down the tall timber?" In Limiting Legislative Terms by Gerald Benjamin and Michael Malbin. Washington, D.C.: Congressional Quarterly Press.
Fett, Patrick, and Daniel Ponder, 1993. "Congressional Term Limits, State Legislative Term Limits, and Congressional Turnover." PS: Political Science and Politics, 26: 211-16.
Guerva, Fernando J., 1991. "Term Limits and Minority Representation in California."
Presented at the Annual Meeting of the American Political Science Association, Washington, D.C.
Hodson, Timothy, 1994. Conventional wisdom or wishful thinking: Staff influence in post
term limit legislatures. APSA Legislative Studies Section Newsletter: Extension of
Remarks 17(2)8-9, 15.
Hodson, Timothy; Rich Jones; Karl Kurtz; and Gary Moncrief, 1995. "Leaders and limits:
changing patterns of state legislative leadership under term limits."
Spectrum: the Journal of State Government 68(3)6-10.
Jones, Charles O. The Presidency in a Separated System. Washington, D.C.: The Brookings
Institutions. 1994.
Karp, Jeffrey A., 1995. "Explaining Public Support for Legislative Term Limits." Public Opinion
Quarterly, Vol. 59, No. 3, pp. 373-391.
Key, V.O., Southern Politics, Chapter 1, "Of the South," 1984.
Kurtz, Karl, 1992. "Limiting terms, what's in store?" State Legislatures, 18:32-34.
-"The Public Standing of the Legislature," National Conference of State Legislatures. Paper presented at the Symposium on the Legislature in the Twenty-
First Century, Williamsburg, Virginia, April, 1990.
Magleby, David D. 1984. Direct Legislation: Voting on Ballot Propositions in the United States. Baltimore, Johns Hopkins University.
Moncreif, Gary F., Joel A. Thompson, Michael Haddon, Robert Hover, 1992. "For
Whom the Bell Tolls: Term Limitations and State Legislatures." Legislative
Studies Quarterly, 17: 37-47.
Neil, T., National Conference of State Legislatures. Denver, Colorado.
Niemi, Richard G., Harold W. Stanley, and Ronald J. Vogel, 1995. "State Economies and State
Taxes: Do Voters Hold Governors Accountable?" American Journal of Political Science, Vol. 39, No. 4, pp. 936-957.
Peterson, Paul E., and Mark Rom, 1984. "American Federalism, Welfare Policy,
and Residential Choices." American Political Science Review, Vo. 83, No. 3. pp. 712-727.
Pierce, Charles, 1992. "The guillotine comes to California: Term limits in the golden state." In Limiting Legislative Terms by Gerald Benjamin and Michael Malbin. Washington, D.C.: Congressional Quarterly Press.
Reed, Robert, and Eric Schausberg, 1994. "An Analysis of the impact of congressional term
limits." Economic Inquiry, 32:79-91.
Rothenberg, Stuart, 1992. How term limits became a national phenomena.
State Legislatures, 18:35-9.
Schmidt, David D., 1989. Citizen Lawmakers: The Ballot Initiative Revolution, Philadelphia:
Temple University Press.
Statistical Abstracts of the United States, 1995, 1996. United States Bureau of the Census.
Washington, D.C.
Will, George, 1992. Restoration: Congress, Term Limits and the Recovery of Deliberative
Democracy. New York: Free Press.
Zisk, Betty, 1987. Money, Media and the Grass Roots: State Ballot Issues and the
Electroal Process. New Bury Park, California: Sage Publishing.
(Note: Data Tables are available from the paper’s author and have not been included online.)