Increasing Diversity or More of the Same? Term Limits and the

Increasing Diversity or More of the Same?

Term Limits and the Representation of Women, Minorities,

and Minority Women in State Legislatures

Susan J. Carroll and Krista Jenkins

Rutgers University

Prepared for delivery at the 2001 Annual Meeting of the American Political

Science Association, San Francisco, California, August 30 - September 2.

The movement to limit the number of terms served by legislators at both national and state levels has

provoked heated debate about the possible effects of term limits. One set of debates focuses on whether term

limits will lead to more diverse legislatures, especially along the lines of race, ethnicity, and gender. Because

term limits will help to break the stranglehold of incumbency by increasing legislative turnover, term-limit

advocates and some scholars have argued that previously underrepresented groups are almost certain to benefit.

For example, term-limits advocate Jonathan Ferry, former Communications Director of the U.S. Term Limits

Foundation, has argued:

Minorities and women have been shut out of electoral politics for too long due to an entrenched

block of white male incumbents. Elimination of this large block of over-represented incumbents

through term limits will create hundreds of open seats in which women and minorities can make

significant gains in representation and create a legislature in which the interests of all groups are

better represented on all issues (1994).

Similarly, Edward H. Crane, President of the Cato Institute, testified before a congressional subcommittee that

women and minorities fare better in open seat races and that term limits would “enhance the competitiveness of

elections and... increase the number and diversity of Americans choosing to run” (1995).

Among scholars, the argument that term limits might well lead to more diverse legislatures by increasing

the representation of previously underrepresented groups has more often been made regarding women than

minorities. For example, R. Darcy, Susan Welch, and Janet Clark have suggested, “Term limitations, when used

as a way to weaken the power of incumbency for example, will speed up the election of women” (1994, 194).

As evidence for this assertion, they examined cohorts of legislators in the lower houses of 21 state legislatures

and found that the most recently elected cohorts had about twice as many women proportionately as did the

more senior cohorts (1994, 146). They concluded, “since women are a much larger proportion of newly elected

legislators, many of the men forced to step down will be replaced by women” (1994, 146). Similarly, Barbara

Burrell has observed, “If terms of office were limited..., as has been adopted in some states and has been

proposed for the U.S. Congress, that should at least in the short run increase the number of women legislators as

it would remove long term incumbents, disproportionately men, from the electoral equation” (1994, 191).

Several other scholars have argued along similar lines that term limits could potentially lead to increases in the

number of women serving in legislatures (e.g., Carroll 1994; Darcy 1992; Carroll and Strimling 1983, 6; Rule

and Norris 1992).

It is more difficult to find scholars who have suggested, as term-limit advocates have, that term limits will

likely lead to increased numbers of minorities serving in legislatures. In general, the literature on minority

politics and representation is far more preoccupied with questions surrounding racial redistricting and the

relative advantages and disadvantages of different electoral systems (e.g., district versus at-large) than with the

potential effects of term limits (see, e.g., McClain and Garcia, 1993; Lublin 1997; Swain 1993; Canon 1999;

Darcy, Hadley, and Kirksey 1993). Nevertheless, in contrast to term-limit advocates, the few scholars who have

speculated about the potential impact of term limits on minorities have generally viewed the likely effects as

either negative or neutral. For example, David A. Bositis, a political scientist and senior research associate at

the Joint Center for Political and Economic Studies, has voiced his concern that term limits would “sweep out

of office the entire cadre of the most experienced, ablest black elected officials, depriving their constituents of

their leadership and hard-won influence” (1992, 7). Unlike Bositis who voiced concerns about the potential loss

of experience and did not directly address the probable effects of term limits on the actual numbers of African

American representatives, W. Robert Reed and D. Eric Schansberg have suggested that, at least in Congress,

term limits would lead to a decline in the numerical representation of minorities because nonwhites have had

longer average tenures than whites (1995, p. 70). Other scholars, focusing on state legislatures, have argued that

term limits are likely to have little impact on the numerical representation of minorities. Joel A. Thompson and

Gary F. Moncrief, in their analysis of retention rates of women and minority state legislators, concluded that

“the number of minority legislators is not likely to be seriously affected in the short run” (1993, 308).

Even in the case of women, where most term-limit advocates and scholars seem to be in agreement that

term limits will create more political opportunities, the view that term limits will lead to increases in numerical

representation has not gone uncontested by either practitioners or scholars. Some activists concerned with

increasing the number of women in public office have argued that term limits are not likely to be an effective

mechanism for increasing women’s representation. Becky Cain, for example, writing as president of the League

of Women Voters of the United States, suggested that campaign finance reform would be a preferable reform to

term limits, allowing women challengers to run competitive races (1994, 48). Among the strongest critics of

term limits have been incumbent women officials who have worried they would be forced out of office without

any guarantee they will be replaced by other women. Former state representative Maxine Berman of Michigan

is one who holds this view. She has expressed her fear that term limits will threaten the incremental gains in

representation which women have made in recent years because women will be forced to leave office at a rate

equal to or greater than the rate at which new women are elected (Hoffman 1998).

Some scholars also have voiced reservations about term limits similar to Berman’s. While acknowledging

that women are likely “to benefit selectively from term limits,” Linda L. Fowler, for example, has argued,

“women are not going to gain by this reform across the board, however, because the more basic problem

remains: too few women run in most states–-in either primaries or open seat races-–to produce significant

increases in the number of female office holders” (1992, 182).

In this paper we examine the question of whether term limits lead to greater diversity among legislators in

terms of their gender, race, and ethnicity. We assess the effect of the implementation of term limits on the

numbers of women, minorities, and minority women serving in state legislatures with data from the 1998 and

2000 elections. Despite the reservations expressed by some practitioners and scholars, the existing literature

leads us to anticipate increases in the number of women serving in legislatures as a result of term limits.

Although the scholarly literature provides less guidance as to what to expect with respect to minorities, we

hypothesize that term limits will result in little change in the numbers of minorities holding state legislative

seats, largely because so many minority legislators are elected from districts with large minority populations,

rather than from predominantly white districts. We expect to find that most minorities who are forced from

office due to term limits are replaced by members of the same racial or ethnic group.

The previous literature on term limits offers even less guidance as to what pattern to expect with regard to

the effect of term limits on the numerical representation of minority women, who are members of both groups

(i.e., women and minorities). However, the work of R. Darcy, Charles D. Hadley, and Jason F. Kirksey and the

interdisciplinary feminist literature are both helpful. Darcy, Hadley, and Kirksey (1993) have provided

compelling evidence that the underrepresentation of African Americans in elective office in the U.S. is due

primarily to the underrepresentation of African American women. Their research has demonstrated that African

American women, like white women, are underrepresented relative to their presence in the population as a

whole while the representation of African American men more closely corresponds to their presence in the

general population. Moreover, similar to the pattern for white women, African American women are

advantaged by multimember districts while African American men are not. These findings suggest that the

pattern for African American women may more closely resemble the general pattern for women than for

African American men. Interdisciplinary feminist writings, however, complicate this expectation. These

writings have suggested that the experiences of African American and other minority women are distinctive

from those of white women and minority men and cannot be separated into gendered and racial/ethnic

components (e.g., Spelman 1988; Crenshaw 1998; Higginbotham 1992; King 1988; Glenn 1992). The feminist

literature thus leads us to anticipate that the effects of term limits on minority women may be unique and that

the patterns we find for minority women may differ from those found for women and for minorities overall.

Description of the Data Set and the Analysis

Although more than a decade has passed since states first began to adopted statutes limiting the number of

terms legislators can serve, the first forced retirements of legislators did not occur until 1996 and then in only

two states, California and Maine (National Conference of State Legislatures, 1999). Several other states joined

California and Maine in implementing term limits for some legislative seats in 1998 and 2000,

and

consequently the 1998 and 2000 elections presented the first opportunity to examine the impact of term limits

on the representation of women and minorities with a data base that extends across several states and more than

one election. In 1998, 203 state legislators in seven states were forced to leave office because of term limits

(Table 1). In 2000, 377 state legislators in 12 states vacated their seats due to term limits (Table 1).

Table 1: Numbers and Proportions of Seats in Various States Which Were Term-Limited and Up for

Election in 1998 and 2000

1998 2000

Senate House Senate House

State N= %= N= %= N= %= N= %=

Arizona -- -- -- -- 7 23.3 15 25.0

Arkansas -- -- 49 49.0 13 76.5 24 24.0

California 11 16 20.0 8 40.0 23 28.8

Colorado 9 18 27.7 11 57.9 10 15.4

Florida -- -- -- -- 11 52.4 54 45.0

Maine 1 11 7.3 7 20.0 16 10.6

Michigan -- -- 64 58.2 -- -- 21 19.1

Missouri

1 -- -- -- -- -- 8 --

Montana -- -- -- -- 13 50.0 34 34.0

Ohio -- -- -- -- 6 37.5 43 43.4

Oregon 2

-- 21 35.0 5 33.3 15 25.0

South Dakota -- -- -- -- 13 37.1 20 28.6

N= 24 179 94 283

We have excluded Missouri from our analysis because term limits will not be implemented across-the-board

for the Missouri House and Senate until 2002; in other words, 2002 will be the first year in which incumbents

who were serving when the term limits measure passed will be forced to retire. However, 1 senator in 1998 and

8 representatives in 2000 became subject to term limits before other members of their respective houses

because they filled mid-term vacancies.

These two senators in Oregon became subject to term limits under exceptional circumstances. They were

originally appointed to mid-term vacancies and consequently became subject to term limits before other members

of the Oregon senate.

For purposes of this analysis, we compiled a data set for all 1998 state house races in the six states that

had by then implemented term limits (Arkansas, California, Colorado, Maine, Michigan, and Oregon) as well as

for all 2000 state house races in these six states and the five states that implemented term limits for the first

time in 2000 (Arizona, Florida, Montana, Ohio, and South Dakota). Each legislative seat was identified as

either term-limited, other open (i.e., no incumbent running for re-election but not term-limited), or not open

(i.e., incumbent seeking re-election).

For each legislative race (seat), we coded additional information about

the sex, race/ethnicity, and party of both pre-election incumbents and candidates who won the general election.

The data were obtained mostly from the Project Vote Smart web site

and the offices of the secretaries of state

in each state.

Information on the race or ethnicity of candidates was not available from the Project Vote Smart web site;

however, we were able to gather information on the race and ethnicity of incumbents and general election

winners from other sources. African Americans serving as state legislators before and after the 1998 and 2000

elections were identified through lists provided by the Joint Center for Political and Economic Studies. Because

no similar clearinghouse exists for information on Latino and Asian American officeholders, we relied on the

clerks of the house and secretaries of the senate in each state for identification of Latino, Asian American, and

Native American pre-election incumbents and post-election winners.

As an additional cross-check on those

whom our sources had not identified as African American, Latino, Asian American, or Native American, we

obtained and visually inspected pictures of legislators over the internet or through legislative directories for

several of the states.

Because of the very small number of cases of minority legislators in term-limited state senate seats both

before and after the 1998 and 2000 general elections,

our analysis is limited to state house seats. Additionally,

because no data are available regarding the race and ethnicity of primary candidates, our analysis is limited to

outcomes of general election races.

The Effect of Term Limits on the Numbers of Women and Minorities in State Houses

Women and minorities fared very differently in races for term-limited seats in the 1998 and 2000 elections

(Tables 2 and 3). Contrary to the expectation of term-limit advocates and many scholars, the number of women

serving in term-limited state house seats actually decreased following each election.

In other words, more

women were forced to vacate seats because of term limits than were elected to seats that were open as a result

of term limits. Across the six states that implemented term limits for state house races in 1998, 47 incumbent

women were forced to leave office as a result of term limits while only 43 women won election to house seats

vacated by term-limited incumbents (Table 2). Similarly, across the 11 states in which term limits were in effect

for house seats in 2000, 70 women who served in term-limited seats left office while only 65 new women were

elected to seats that were open as a result of term limits (Table 3).

Table 2: Numbers and Proportions of Term-Limited, Other Open, and Non-open Seats in State

Houses Represented by Members of the Same Group Before and After the 1998 Elections in

Six States

Type of Seat Pre-

Election

Post-

Election

Women Term-Limited 47 26.3 43 24.0

Other Open 11 25.0 10 22.7

Not Open 87 25.4 92 26.8

Minorities Term-Limited 22 12.3 29 16.2

Other Open 3 6.8 4 9.1

Not Open 28 8.1 30 8.7

African Americans Term-Limited 12 6.7 15 8.4

Other Open 1 2.3 2 4.5

Not Open 18 5.2 18 5.2

Minority Women Term-Limited 11 6.1 9 5.1

Other Open 2 4.5 1 2.3

Not Open 7 2.0 8 2.3

African American

Women

Term-Limited 7 3.9 3 1.7

Other Open 0 0.0 1 2.3

Not Open 4 1.2 5 1.5

States included are Arkansas, California, Colorado, Maine, Michigan, and Oregon. There were no minorities in term-

limited seats in Maine and no African Americans in California, Colorado, and Maine either before or after the

election. There were no minority women in term-limited seats in Maine either before or after the election. The only

states with African American women in term-limited seats before or after the election were Arkansas, Michigan, and

Oregon.

Table 3: Numbers and Proportions of Term-Limited, Other Open, and Non-open Seats in State

Houses Represented by Members of the Same Group Before and After the 2000 Elections in

11 States

Type of Seat Pre-

Election

Post-

Election

Women Term-Limited 70 25.5 65 23.6

Other Open 25 27.2 31 33.7

Not Open 170 26.2 175 27.0

Minorities Term-Limited 32 11.6 39 14.2

Other Open 12 13.0 14 15.2

Not Open 74 11.4 75 11.6

African Americans Term-Limited 22 8.0 24 8.7

Other Open 5 5.4 5 5.4

Not Open 36 5.6 38 5.9

Minority Women Term-Limited 8 2.9 11 4.0

Other Open 4 4.3 5 5.4

Not Open 22 3.4 26 4.0

African American

Women

Term-Limited 7 2.5 8 2.9

Other Open 0 0.0 1 1.1

Not Open 11 1.7 14 2.2

States included are Arizona, Arkansas, California, Colorado, Florida, Maine, Michigan, Montana, Ohio, Oregon, and

South Dakota. There were no minorities in term-limited seats in Maine or Oregon and no African Americans in

California, Colorado, Maine, Montana, Oregon, or South Dakota either before or after the election. The only states

with minority women in term-limited seats before or after the election were Arizona, Arkansas, California, Florida,

Michigan, and Ohio. The only states with African American women in term-limited seats before or after the election

were Arizona, Arkansas, Florida, Michigan, and Ohio.

Although the number of women serving in term-limited seats decreased in each election, women were able

to compensate for these losses by gains made in races where they challenged incumbents (in both 1998 and

2000) and where seats were open for reasons other than term limits (in 2000) (Tables 2 and 3). As a result, the

total number of women serving in the lower houses of the six states that implemented term limits in 1998

remained the same--145--before and after the 1998 elections, and the total number of women state

representatives in the 11 states affected by term limits in 2000 actually increased following the 2000 elections

from 265 to 271.

The numbers of women elected to term-limited seats did vary notably across the six states. In only two

(Michigan and Oregon) of the six states which implemented term limits in 1998 were more women elected to

seats that were open due to term limits than were forced out of office because of term limits.

In one state

(California) the numbers of women serving in term-limited seats before and after the 1998 elections stayed the

same, and in three states (Arkansas, Colorado, and Maine), the numbers decreased.

Similarly, in only three

(California, Ohio, and South Dakota) of the 11 states affected by term limits in 2000 were more women were

elected to house seats which were open as a result of term limits than were forced out of office because of term

limits.

In three states (Arizona, Florida, and Oregon) the numbers of women in term-limited seats stayed the

same while women’s numbers decreased in five states (Arkansas, Colorado, Maine, Michigan, and Montana).

In contrast to the pattern for women state representatives, minorities increased their representation in term-

limited seats in both the 1998 and 2000 elections. Overall, the number of minorities serving in term-limited

house seats increased from 22 to 29 following the 1998 elections (Table 2) and from 32 to 39 following the

2000 elections (Table 3). In 1998 the most notable increases came in Michigan, where the number of minorities

serving in term-limited seats increased from five to eight, and in Colorado, where the number increased from

two to four. In 2000 the largest increases were again in Michigan, where the number of minority

representatives in term-limited seats increased from three to six, and in Florida, the state with by far the largest

number of minorities serving in term-limited seats, where the number increased from 16 to 18. To some extent,

gains made by minorities in term-limited seats seem to have been part of a larger pattern of gains for minorities

more generally; in both elections minority representatives also increased their numbers slightly in other open

seat races and in races where minority candidates challenged incumbents (Tables 2 and 3). Nevertheless, the

gains for minorities in races for term-limited seats were proportionately greater than in races for other types of

seats.

Just as minorities in general increased their numerical representation in term-limited seats, so too did

African Americans (Tables 2 and 3). Increases in 1998 were concentrated in the states of Arkansas and

Michigan; in each state African Americans increased their numbers by two in races where incumbents had been

forced out by term limits.

In 2000, the only increase for African Americans was again in Michigan, where six

African Americans were elected to seats open because of term limits while only three African Americans were

forced out of office by term limits.

Table 4: Proportions of Term-Limited State House Seats Previously Held by Various Groups Where a

Member of the Same Group Won the 1998 General Election

% of Term-Limited House Seats

Held by a Group Member Where a

Member of the Same Group Won

the General Election

Women 25.5

Minorities 81.8

African Americans 91.7

Minority Women 27.2

African American Women 14.3

States included are Arkansas, California, Colorado, Maine, Michigan and Oregon.

N= 47 for women, 22 for minorities, 12 for African Americans, 11 for minority women, and 7 for

African American women.

An examination of the rates at which term-limited incumbents were replaced by members of the same

group (e.g., women, minorities, and African Americans) provides a partial explanation for why minorities

increased their numbers in term-limited seats while the numbers of women in these seats declined. In contrast to

the pattern for women, minorities in term-limited seats were usually replaced by other minorities; this was

particularly true for African Americans (Tables 4 and 5). The high level of success which minorities had in

retaining seats is undoubtedly a reflection of the fact that many minority legislators represent majority-minority

districts. Nevertheless, the high rate of retention of term-limited seats gave minorities a strong base on which

they could build numerical increases by taking advantage of a few of the opportunities presented by seats

vacated by term-limited, non-minority incumbents. In contrast, because term-limited women incumbents were

replaced by other women in only about one-fourth of all cases in both elections, women had a proportionately

smaller base on which to build and would have had to win many term-limited seats previously held by men in

order simply to maintain their numbers.

Table 5: Proportions of Term-Limited State House Seats Previously Held by Various Groups Where a

Member of the Same Group Won the 2000 General Election

% of Term-Limited House Seats

Held by a Group Member Where a

Member of the Same Group Won

the General Election

Women 28.6

Minorities 90.6

African Americans 95.5

Minority Women 12.5

African American Women 14.3

States included are Arizona, Arkansas, California, Colorado, Florida, Maine, Michigan, Montana,

Ohio, Oregon, and South Dakota.

N= 70 for women, 32 for minorities, 22 for African Americans, 8 for minority women, and 7 for

African American women.

Another part of the explanation for why minorities increased their numbers in term-limited seats emerges

from an examination of the gains they made in term-limited seats previously held by non-minorities. In the 1998

elections a total of 11 term-limited seats (in four states) held by whites prior to the election were captured by

minorities. Of these 11 seats, four seats, two in Arkansas and two in Michigan, were won by African Americans;

all four seats were in majority-black districts where a white incumbent had held office for 14 to 20 years. In 2000,

10 term-limited seats (in six states) held by whites prior to the election were won by minorities. Of these 10 seats,

three seats, all in Michigan, were won by African Americans. One of these seats was in a majority-minority

district (43 percent black and 12 percent Latino) which the white incumbent had represented for the past six years

while the other two were in districts where majorities of the voters were white.

Thus, it appears that the implementation of term limits enabled African American candidates to capture

some seats in districts where a majority or near-majority of voters were African American, but where the power

of incumbency had helped white representatives remain in office for a long period of time. African Americans

also gained a couple of seats in majority-white districts.

For the one Asian American in 1998, the one Native American in 2000, the six Latinos in 1998, and the

six Latinos in 2000 who were able to capture term-limited seats previously held by whites, the picture is

somewhat different. The Asian American and Native American were both elected in districts that were

predominantly white (although a reservation is partially located in the district of the Native American). While

two Latinos in 1998 and two in 2000 were elected in districts with substantial Latino populations (i.e., greater

than 40 percent), nevertheless a majority of Latinos were elected in districts with predominantly white

populations. One Latino in 1998 was elected in a district where only one-fifth of the voters were Latino, and

three Latinos in 1998 and four in 2000 were elected in districts where Latinos were less than 12 percent of the

population.

Thus, term limits seem to have presented two different type of opportunities for minorities, one of which

more often benefitted African Americans while the other more often worked to the advantage of Latinos (and to

a lesser extent Asian Americans and Native Americans). By removing from office some white incumbents who

had long served in majority-minority districts, term limits provided opportunities for minorities, mostly African

Americans, to capture seats in those districts. In contrast, several minorities, especially Latinos, were successful

in winning seats in districts where whites were a majority, often a sizable majority, of the population. It is the

combination of these different patterns, one more common for African Americans and one more common for

Latinos, that helps to account for the post-1998 increase in numerical representation of minorities in term-

limited seats.

The Effects of Term Limits on the Numbers of Minority Women

Minority women are included in the above analysis of “women” as well as in the analysis of “minorities.”

Yet, women and minorities fared very differently in races for term-limited house seats in the 1998 and 2000

state legislative elections, with minorities increasing their numerical representation in term-limited seats at the

same time that the numbers of women in term-limited seats decreased. Although analyzing women of color

separately is intrinsically important, as we argued earlier, since their life experiences and identities are in many

ways distinct from those of both white women and men of color, the fact that women and minorities fared

differently overall in elections for term-limited state house seats in 1998 and 2000 makes all the more intriguing

the question: how did minority women fare?

For the 1998 elections the pattern for minority women more closely resembles the pattern for women than

it does for minorities (Table 2). While the number of cases is small, the number of minority women serving in

term-limited state house seats decreased by two following the 1998 elections. Moreover, the decline in numbers

was concentrated among African American women (Table 2). While African American women were able to

increase their numbers by one in seats open for reasons other than term limits and also by one in races where

they challenged incumbents, the number of African American women serving in term-limited seats decreased

by four. In contrast, Latinas actually increased their representation in term-limited seats from four to six.

(There were no Asian American women or Native American women serving in term-limited state house seats

either before or after the 1998 elections.)

Unlike 1998 where the pattern for minority women more closely resembles the pattern for women overall,

in the 2000 elections the pattern for minority women parallels the pattern for minorities more generally (Table

3). While the number of cases is again small, the number of minority women serving in term-limited seats

increased by three following the 2000 elections. Like minorities overall, minority women also increased their

numbers slightly in seats open for reasons other than term limits and in seats where an incumbent sought re-

election (Table 3). Part of the increase in the numbers of minority women in term-limited seats was due to the

victories of African American women, who won one more term-limited seat than they vacated. Latinas also

increased their numbers in term-limited seats from one to three.

Like the pattern for women and unlike the pattern for minorities, few minority women in term-limited state

house seats were replaced by women who shared their race or ethnicity in either election (Tables 4 and 5). This

was true for minority women in general and for African American women in particular (Tables 4 and 5). Of the

seven African American women who were forced out of office by term limits in 1998, only one (in Michigan)

was replaced by another African American woman. Five of the seven African American women incumbents

subject to term limits in 1998 served in the Arkansas state house, and all were replaced by African American

men. The lone African American woman who was term-limited in Oregon in 1998 was replaced by a white

woman.

Of the seven African American women legislators who were forced to leave office in 2000 due to term

limits, five (including four in Florida) were replaced by African American males and another was replaced by a

Latino. Only one African American woman, an Ohio legislator, was replaced by an African American woman.

Term-limited Latinas were somewhat more likely to be replaced by another Latina in 1998. Of the four

Latinas forced out of office by term limits, two (one in California and one in Michigan) were replaced by

Latinas. The other two Latinas were replaced by Latinos. The sole Latina who was forced by term limits to

leave office in 2000 also was replaced by a Latino.

Discussion and Conclusions

Findings from our analysis of electoral outcomes in states where term limits were in effect in 1998 and

2000 suggest that the answer to the question of whether term limits lead to more diverse legislatures is not

straightforward. Moreover, our findings are sometimes contrary to the expectations of activists and scholars.

To date, women, minorities, and minority women have fared differently under term limits. Contrary to the

expectations of term-limit advocates and many scholars of women and politics, more women state

representatives were forced out of office by term limits in both 1998 and 2000 than were elected to seats

vacated by term-limited incumbents.

In contrast to this pattern for women, and contrary to the expectations of

some scholars who suggested that minority representation would likely stay the same or decrease under term

limits, the number of minorities in term-limited state house seats increased following both elections. Minority

women fit neither the pattern for women nor for minorities. While the number of minority women (African

American women in particular) serving in term-limited state house seats decreased slightly following the 1998

elections, the number of minority women (and African American women, more specifically) in term-limited

seats increased slightly following the 2000 elections. (The number of Latinas increased slightly in both

elections.)

Although minorities increased their representation in term-limited seats consistent with the projections of

some term-limit proponents, a deconstruction of the category “minorities” reveals a more complex underlying

set of patterns. African Americans increased their representation largely because they were able to win seats in

majority-minority districts which were held by white incumbents who had served long tenures; however, they

won fewer seats in majority-white districts. This raises the question of whether gains for African Americans

under term limits will continue once all existing majority-minority districts have been subject to term limits.

The implications of our findings are somewhat more promising for Latinos who were able to win several

term-limited seats in districts where Latinos were a minority (and even a small minority). The findings suggest

that Latinos may continue to benefit from term limits if strong Latino candidates who can appeal to primarily

white constituencies continue to come forward.

Although the numbers were small, the findings of decreases in the number of African American women in

term-limited seats in 1998, contrary to the pattern for African American men, suggests that the implementation

of term limits may have a somewhat different impact on the representation of African American women and

men. Despite the fact that African American women increased their numbers in term-limited seats by one in

2000, the paucity of successful African American women candidates for term-limited seats in 1998 is troubling

for anyone who might be concerned with increasing the numbers of minority women officeholders.

The mere existence of more political opportunities has not been sufficient to increase the number of

women serving in office. The concerns of incumbent women public officials, who fear they will not be replaced

by other women when they are forced to leave office as a result of term limits, seem to have merit. Large

proportions of women legislators who relinquished their seats in 1998 and 2000 were replaced by men. This is

very different than the pattern for minorities, the vast majority of whom were replaced by members of the same

racial or ethnic group. Women legislators who know in advance that they will be forced to leave office by term

limits could help increase the numbers of women in state legislatures by working to insure that they are

replaced by other women, just as their minority colleagues are usually replaced by other minorities. This applies

to minority women as well as to white women since most of the minority women who were forced to vacate

their seats in 1998 and 2000 were replaced by men. Efforts to recruit women candidates to run for the seats of

term-limited women legislators, as well as for the seats of other term-limited legislators, could help insure that

the opportunities offered by term limits are eventually translated into the reality of increases in the number of

women legislators.

Additional research is certainly needed as term limits are implemented in more states in future elections.

However, thus far the answer to the question of whether term limits are leading to more diverse legislatures

clearly depends upon the group being considered. To date, minorities have fared better than women under term

limits, and for minority women the picture is decidedly mixed.

1. Term limits generally are phased in over multiple election cycles so that the entire legislature does not turn over at the

same time. Consequently, only a portion of the legislators in either chamber of any of the affected states was subject to

term limits in each of the years 1996, 1998, or 2000.

2. Our numbers are based on information provided both by the National Conference of State Legislatures and by the offices

of each state’s secretary of state. We have excluded eight representatives from Missouri in 2000 from the analysis in this

paper because of the exceptional circumstances under which they became subject to term limits before term limits were

implemented more broadly for their legislative bodies. See Table 1 for more information.

3. All term-limited seats in this analysis were in single-member districts except for term-limited house seats in Arizona and

South Dakota in 2000, where two representatives were elected from each district. For districts in these two states, we coded

two races per district but included all new candidates running in the district as possible candidates for each seat. Thus, if a

district had one seat where an incumbent was seeking re-election and another which was open due to term limits, we coded

the two races as separate cases, one as an incumbent-occupied seat and one as a term-limited seat with all candidates in the

race counted as candidates for the incumbent-occupied seat and all candidates except for the incumbent counted as

candidates for the term-limited seat.

4. Project Vote Smart is the major program of The Center for National Independence in Politics, a national non-partisan

501(c)(3) organization focused on providing citizens/voters with information about the political system, issues, candidates,

and elected officials. Vote Smart collects demographic data on candidates in statewide and state legislative races across the

country. Additionally, Vote Smart administers a “National Political Awareness Test” which measures candidates’ stances

on issues of importance in each state. Vote Smart has collected independent, factual information on over 13,000 candidates

and elected officials. Data from Vote Smart were obtained from their web site (www.vote-smart.org).

5. While we recognize the multiplicity of racial and ethnic categories, the few verifiable sources for race and ethnicity

information on state legislators prevented us from going beyond standard identifications of individuals as either white,

African American, Latino, Asian American, or Native American.

6. N=2 before and N=6 after the 1998 elections; N=10 before and N=10 after the 2000 elections.

7. For a much more detailed analysis of how women candidates for state legislative seats fared under term limits in the

1998 and 2000 elections, see Carroll and Jenkins forthcoming.

8. Although the relationship between gender and pre- and post-election occupancy of term-limited seats is weak statistically

(Yules Q= -.06 for 1998 and -.05 for 2000), the decreases in the number of women serving in these seats, while small, are

nevertheless politically significant since they are contrary to the expectations of the literature which suggests a moderate,

positive relationship between gender and pre- and post-election occupancy of term-limited seats.

9. The number of women serving in term-limited seats increased from 16 to 17 in Michigan and from five to nine in

Oregon.

10. In Arkansas nine women were forced to vacate term-limited seats, and only eight women were elected to seats that

opened up as a result of term limits. Similarly, the number of women serving in term-limited seats declined from nine to

two in Colorado and from two to one in Maine.

11. The number of women serving in term-limited seats increased from five to eight in California, from 12 to 14 in Ohio,

and from three to five in South Dakota.

12. The number of women serving in term-limited seats declined from six to one in Arkansas, from three to one in

Colorado, from five to four in Maine, from seven to four in Michigan, and from seven to six in Montana.

13. In contrast, the number of African Americans serving in term-limited state house seats in Oregon, the only other state

with an African American incumbent who was forced out of office by term limits, declined from one to zero following the

election.

14. Complicating the picture even more, this is not true for state senates where the number of women serving in term-

limited seats increased in 1998 and stayed the same in 2000. See Carroll and Jenkins forthcoming.

Notes

References

Bositis, David A. “Will Term Limits Help?” Focus. January, 1972: 7.

Burrell, Barbara. 1994. A Woman’s Place Is in the House: Campaigning for Congress in the Feminist Era. Ann Arbor:

Michigan.

Cain, Becky. 1994. “Term Limits: Not the Answer to What Ails Politics.” In The Politics and Law of Term Limits, eds.

Edward H. Crane and Roger Pilon. Washington, D.C.: Cato Institute.

Cannon, David T. 1999. Race, Redistricting, and Representation. Chicago: University of Chicago Press.

Carroll, Susan J. 1994. Women as Candidates in American Politics. 2

ed. Bloomington: Indiana University Press.

Carroll, Susan J., and Wendy S. Strimling. 1983. Women’s Routes to Public Office: A Comparison with Men’s. New

Brunswick, NJ: Center for the American Woman and Politics.

Carroll, Susan J., and Krista Jenkins. Forthcoming. “Unrealized Opportunity? Term Limits and the Representation of Women

in State Legislatures.” Women & Politics.

Crane, Edward H. “Testimony before the Subcommittee on the Constitution, Committee on the Judiciary, U.S. Senate.”

January 25, 1995. http://www.cato.org/testimony/ct-cr-1.html (March 7, 1999).

Crenshaw, Kimberle. 1998. “Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of

Antidiscrimination Doctrine, Feminist Theory, and Antiracist Politics,” in Feminism and Politics, ed. Anne

Phillips. Oxford: Oxford University Press.

Darcy, R., Susan Welch, and Janet Clark. 1994. Women, Elections, and Representation. 2

ed. Lincoln: University of

Nebraska Press.

Darcy, R. 1992. “Electoral Barriers to Women.” In United States Electoral Systems: Their Impact on Women and Minorities,

eds. Wilma Rule and Joseph F. Zimmerman. New York: Praeger.

Darcy, R., Charles D. Hadley, and Jason F. Kirksey. 1993. “Election Systems and the Representation of Black Women in

American State Legislatures.” Women & Politics 13: 73-89.

Ferry, Jonathan. “Women, Minorities and Term Limits: America’s Path to a Representative Congress.” U.S. Term Limits

Foundation Outlook Series. July 1994. http://www.teleport.com/~prf/tl/women.shtml (March 7, 1999).

Fowler, Linda L. 1992. “A Comment on Competition and Careers.” In Limiting Legislative Terms, eds. Gerald Benjamin and

Michael J. Malbin. Washington, D.C.: CQ Press.

Glenn, Evelyn Nakano. 1992. “From Servitude to Service Work: Historical Continuities in the Racial Division of Paid

Reproductive Labor.” Signs 18:1-43.

Higginbotham, Evelyn Brooks. 1992. "African-American Women's History and the Metalanguage of Race." Signs 17:251-74.

Hoffman, Kathy Barks. “Term Limits May Have Kept Women from Increasing Numbers in State House.” Associated Press.

November 7, 1998.

King, Deborah K. 1988. “Multiple Jeopardy, Multiple Consciousness: The Context of a Black Feminist Ideology.” Signs

14:42-72.

Lublin, David Ian. 1997. The Paradox of Representation: Racial Gerrymander and Minority Interests in Congress. Princeton:

Princeton University Press.

McClain, Paula D., and John A. Garcia. 1993. “Expanding Disciplinary Boundaries: Black, Latino, and Racial Minority

Groups in Political Science.” In Political Science: The State of the Discipline II, ed. Ada W. Finifter. Washington, D.C.:

American Political Science Association.

National Conference of State Legislatures. “Term Limits for State Elected Officials.” February 25, 1999.

http://www.ncsl.org/programs/legman/about/termlim.htm#lifetime (March 13, 1999).

Reed, W. Robert, and D. Eric Schansberg. 1995. “The House under Term Limits: What Would It Look Like?” Social Science

Quarterly 76(4): 698-719.

Rule, Wilma, and Pippa Norris. 1992. “Anglo and Minority Women’s Underrepresentation in Congress: Is the Electoral

System the Culprit?” In United States Electoral Systems: Their Impact on Women and Minorities, eds. Wilma Rule and

Joseph F. Zimmerman. New York: Praeger.

Spelman, Elizabeth V. 1988. Inessential Woman: Problems of Exclusion in Feminist Thought. Boston: Beacon.

Swain, Carol M. 1993. Black Faces, Black Interests: The Representation of African Americans in Congress. Cambridge:

Harvard University Press.

Thompson, Joel A., and Gary F. Moncrief. 1993. “The Implication of Term Limits for Women and Minorities: Some Evidence

from the States.” Social Science Quarterly 74(2): 300-309.