How good is the West Virginia state legislature? How capable is it of doing its job? The capacity of this pivotal institution affects the quality of governance in the state, and the quality of life for its citizens. It is therefore in the state's best interests to evaluate the legislature periodically. This evaluation is difficult, however, because a legislature's job is multidimensional, and few clear standards of quality exist. But the state legislature's importance, and the ability of the political system to make changes to improve it, make the difficult task of evaluation worthwhile.

There have been many evaluations of U.S. state legislatures throughout American history. Some of these evaluations have been rigorous and scholarly, others have been impressionistic and unsystematic. Indeed, a spate of reasonably thorough and negative evaluations by academics and journalists in the 1960s and early 1970s helped spark a concerted effort to improve state legislatures nationwide (Heard 1966; Burns 1971; Herzberg and Rosenthal 1972). At that time, legislatures were accused of being unprofessional, lacking in the institutional capacity needed to deal with the problems facing modern states. Holding the U.S. Congress as a model, critics contended that state legislatures had too few staff, too little time, and not enough legislative experience to do their increasingly important and difficult job. Ironically, after almost 30 years of state legislatures heeding this advice and beefing up their institutional capacity, there is a lively debate today as to whether these changes were positive or negative. Once called "the most extreme example of institutional lag" in the country (Heard 1966), we are now cautioned about what is derisively called the "congressionalization" of the state legislature (Rosenthal 1989).

With competition over limited resources, states choose to develop their legislative capacity to a greater or lesser degree. However, there is disagreement over how to measure legislative capacity. Some scholars have measured the legislative capacity of state legislatures by comparing them to the U.S. Congress or an abstract "legislative capacity" scale (Grumm 1971; Squire 1992). Although this may be useful for ranking the legislative capacity of all state legislatures against one another, it is more appropriate to compare state legislatures with their peer institutions. This gives a more practical impression of a state legislature's capacity, and allows for a discussion of realistic improvements that can be made.

Some scholars have measured state legislative capacity as the level of resources the legislature has at its disposal (Grumm 1971; Knight 1988; Squire'1992). The more resources, the argument goes, the better the legislature. This is often translated into a few easily measured characteristics that are based primarily on a state's ability and willingness to spend money on its legislature, such as legislative pay, legislative staffing, and the legislature's total annual budget. While these kinds of measures certainly tap a dimension of legislative capability, focusing on them exclusively is misleading. Money-based scales are biased against smaller, poorer states, such as West Virginia. It therefore should be no surprise that states like West Virginia, Wyoming, North Dakota, and Arkansas consistently rank at the bottom of these scales, while large and wealthy states like California, New York, Massachusetts, and Pennsylvania rank at the top.

Legislative capacity is a multidimensional concept. While money and the things a legislature can purchase with it are undeniably important, legislatures have other attributes that cost no tax dollars and can be used by small states to improve their legislative capacity. For example, a legislature can improve its legislative capacity by efficiently organizing its committee system, setting up rules that facilitate decision-making, empowering its oversight committees, and so forth. Interestingly, the measures of legislative capability which rely most on monetary measures are highly correlated with a state's urbanization and population level, while the measures which have defined the concept more broadly are less highly correlated with these characteristics (Grumm 1971; LeLoup 1978). And perhaps most importantly, the broad-based scales had low correlations with the money-based scales developed in comparable time periods, indicating that focusing on monetary resources alone is misleading (Burns 1971; Bowman and Kearney 1988).

This article evaluates the West Virginia state legislature on a variety of criteria selected to tap the many dimensions of legislative capacity. These criteria are grouped according to the state legislature's three central functions: law-making, oversight of the bureaucracy, and representation.

Law-Making

The state legislature's central job is to solve public problems through the development of law. The West Virginia legislature's law-making capacity is measured by carefully assessing its available resources, the quality of its committee system, and its overall legislation processing performance.

Law-Making Resources

Most of the resources used in making law involve money, either directly or indirectly. Because of this, poor states and sparsely populated states are at a comparative disadvantage. West Virginia has a double disadvantage because it is both poor and sparsely populated. Considering these disadvantages, it is surprising that West Virginia provides its state legislature a level of resources that is reasonably close to the national average. For example, the median state legislature spent about $13.7 million on itself in 1988 on such items as pay and benefits for legislators and their staff, office supplies, building maintenance, and so forth. The West Virginia legislature spent about $9.6 million to facilitate its operations, placing it 34th among the states (U.S. Burean of the Census 1990). While West Virginia spent less than the average amount, considering it ranks 34th in population and 49th in per capita income, this ranking on legislative expenditures shows a strong concern for the legislature in West Virginia as compared to other states.

On a per capita basis, West Virginia actually spent more than the average state. In 1988,1 each West Virginian paid $5.13 toward the cost of the state legislature, versus $4.38 for a citizen in the median state. Of course, larger states can take advantage of economies of scale and provide their citizens with more legislative resources for less money per taxpayer. However, on a per capita income basis, West Virginia puts far more effort (in terms of its available resources) into funding the state legislature than is the norm in the U.S. For example, it Californians each spent as much of their income on their legislature as West Virginians do, they would have spent over $265 million in 1988, instead of the $172 million they actually spent.

But effort and results are not the same thing, and try as it might, West Virginia consistently provides its legislature with fewer resources than the typical state legislature. Legislative staffing provides a good example. Staff are a vital resource, allowing legislators to investigate the public problems facing the state and to evaluate the competing claims of advocates in an impartial manner. Increased staffing was one of the principal reforms advocated by academics, journalists and others prior to the nationwide push toward state legislative reform in the 1960s and 1970s (Burns 1971). Nationwide, state legislative staff has increased dramatically in the last 30 years, with a 24% increase in their numbers between 1979 and 1988 alone (Weberg 1988). But competent staff cost a great deal of money, and many smaller states like West Virginia have not increased their staffs as quickly as some other legislatures.

The West Virginia legislature has addressed the problem of the need for more staff and the lack of money to pay for them by relying heavily on staff who only work for the legislature during the legislative session. While West Virginia ranks 37th in the level of

I This is the last year for which comparative data is available,

year-round staff, it ranks 11 th in the use of session-only staff (National Conference of State Legislatures 1988). Overall, West Virginia ranks 30th in the nation in total staff support, with 381 year-round and session-only staff members in 1988, the last year for which such information is available.

While session-only staff are an economical way to address the pressing need for assistance legislators have during the hectic legislative session, relying on them heavily has its drawbacks .2 Those who only work at a job for two or three months out of the year are amateurs by definition, and are less knowledgeable in the substance and politics of public policy than those who devote their full attention to it. Also, fresh recruiting every year probably leads to less continuity than when staff have permanent jobs in the legislature. Many people will likely move on to more secure positions to which they can devote all of their energies. And finally, there is the concern that the pool of available persons who have schedules flexible enough to work for only two months at a time (and who have the incentive to do so) is dominated by lawyers and others who work for groups with a direct interest in pending legislation. If this is the case, as it has been argued it is in West Virginia, the independence that increased staffing is supposed to provide legislatures from interest groups is negated (Oxendale and Hammock n.d.) '

West Virginia lags even further behind the average state in the area of legislators' compensation than in either staffing or in total legislative expenditures. While few people enter the state legislature with the expectation of making a lot of money at it, the level of pay legislators receive affects the legislature's law-making capacity. If the pay is so low that it causes members great financial hardship to serve, they will leave the legislature sooner than otherwise, and the legislature will have a difficult time recruiting high quality people to replace them. This results in a lack of experience among members. Further, the more time a legislator has to devote to his/her "other job" in order to make ends meet, the less time he/she has to study public problems and propose solutions to them. Of course, the "citizen legislature," in which a variety of average citizens serve part-time, has been valued highly in this country since the time of Thomas Jefferson, and only a few states pay their legislators enough so that they can be full-time representatives. However, compensation levels are an important influence on how much time and effort a member can devote to the legislature.

Salary paid to legislators for the two-year period 1990-91 ranged from $200 in New Hampshire to $115,000 in New York, with a median of $30,000 (Council of State Governments 1992). West Virginia legislators earned $13,000 in this period, ranking 42nd lowest in the nation. Even with major pay raises between 1969 and 1985, West Virginia legislators still work hard for very little money. Again, considering West Virginia's low ranking in per capita income and population, this ranking for legislative compensation is not unexpected. However, while this low level of compensation is understandable, it does not reduce the financial hardship on West Virginia state legislators.

As alluded to earlier, one possible result of the frustration of low pay and little staff assistance is a higher rate of voluntary turnover among legislators (Rosenthal 1974b; Squire 1988). That is, as the job is more difficult and the compensation remains low, more members will choose not to run for reelection. This reduces another key legislative resource: legislative experience. Those who have served several terms, who have seen the same issues arise time and again, and who have learned the ins and outs of the complicated legislative process bring to the legislature a wealth of experience that can make legislative deliberation more thorough, more independent, and more efficient. If a large number of representatives are new each session, and if legislators do not devote at least several terms each to law-making, the legislature is a weaker institution as compared to those other political institutions that do have a substantial number of members with long experience, such as the bureaucracy and interest groups.

In 1990, West Virginians elected a House of Delegates with almost one-third of its members having never served there before (Kopp 1991). Further, almost one-third of the state's senate seats up for election in 1990 were captured by those with experience in neither the House nor the Senate. This was the second highest rate of turnover in the nation that year (Council of State Governments 1992). The story was the same in 1988, when 41% of the legislature was new and West Virginia had the highest rate of turnover in the nation (Council of State Governments 1990). Due to the coincidence of the West Virginia legislature having less turnover than usual in the election of 1992 ("only" 26.1%), and turnover nationwide being much greater then usual, the most recent election placed West Virginia as the median state in terms of turnover (Van Son 1993). While it is not entirely clear why turnover is so high in the West Virginia state legislature, its effect is a low level of legislative experience. Perhaps this high level of turnover has dampened the interest in legislative term limitations that has been seen in other states.

The final legislative resource to be discussed is the length of the legislative session. When the session is short, there is little time to consider fully proposed legislation. Further, many states work under a constitutional limit on the length of the session. Often these restrictions were set many years ago when states had fewer and simpler problems. A constitutional limit takes away from the legislature the decision of how much time is required to address adequately the problems at hand. Indeed, the extension of the length of legislative sessions and the removal of these constitutional limits were high on reformers' lists of ways to improve state legislatures in the 1960s and 1970s (Burns 1971).

In the last 30 years, many states have increased the maximum length of their legislatures' sessions, shifted to annual from biennial sessions, and/or removed constitutional limits on sessions. In 1973, West Virginia changed from a 60-day regular session in odd years and a 30-day budget session in even years to a 60-day annual regular session, thus increasing the total number of meeting days by one-third (Holmes 1989). But as with 29 other states, the West Virginia legislature continues to labor under a constitutional restriction on its session length (Council of State Governments 1990). Citizens suspicious of encroaching government power make the removal of these provisions (which usually require a referendum) very difficult. For example, three times in the last 25 years, Kentucky voters have rejected a proposal to move from biennial to annual legislative sessions (Jewell and Miller 1988).

Despite this constitutional limit on its sessions, West Virginia's legislature met in 1990 and 1991 for a total of 174 calendar days (including special sessions), which was the 28th longest in the country. Because it costs the state nearly $30,000 for every day the legislature meets (for staff salaries, building maintenance, expenses for legislators, etc.), it is perhaps surprising that West Virginia can afford to rank so high here. And perhaps there is less need for an extended session in a state like West Virginia with a small, rural, homogeneous population than, for example, a state with a large, diverse, and urban population like New York. This is not to say that the West Virginia legislature has an overabundance of time to consider the issues before it; anyone observing the hectic last week of a legislative session will know better. But it appears that West Virginia's session length is perhaps reasonably adequate as compared to other states.

The Committee System

In U.S. legislatures at both the state and national levels, most law-making gets done in standing committees. Legislatures divide themselves into smaller groups to allow for a careful consideration of each proposal by a few of its members, rather than settle for a cursory inspection of each bill by the legislature as whole. This delegation of legislative authority to committees encourages greater specialization among legislators and the more efficient and thorough processing of legislation.

State legislatures vary dramatically in the number of committees into which they divide themselves, from 73 in the Missouri legislature, to only 10 in Maine. State legislatures have traditionally shown no disinclination to form committees; in fact, the principal problem reformers of the 1960s and 1970s saw in this regard was that legislatures formed too many committees (Burns 1971). When there are too many committees, the legislature is a confused place, with competition for members' time and attention taking energy away from law-making. The West Virginia legislature reduced the number of committees in both chambers between 1955 and 1969 from 54 to 28 and has since kept the total number of committees at about that level (Reeves 1985). Even after 30 years of committee consolidation in legislatures across the country, West Virginia's 30 legislative committees in 1989 compares favorably to the U.S. median of 37 (Council of State Governments 1992).

Another aspect of a committee system that is related to the number of committees is the number of committee assignments each legislator has. The fewer committees to which a member is assigned, the more he/she can focus his/her attention. This may be as simple as being able to attend all of a committee's hearings, rather than being forced to choose between two hearings scheduled for different committees at the same time. If a committee is to fulfill its information gathering and deliberative functions, its members must have a minimum amount of competing claims on their time.

In the West Virginia state Senate, members are extremely busy, with an average of 5.7 committee assignments in the 1991-92 session (Kopp 1991). This is much higher than the U.S. average of 3.2 assignments (American Society of Legislative Clerks and Secretaries 1988). West Virginia delegates have a much greater opportunity to specialize in their commit

tee work than state senators because they average only 2.9 committees each, which is quite comparable to the U.S. average of 3.0 committee assignments per House member.

While it is traditional for a state's Senate to be less specialized in its legislative activities than its House, taking a broader view of policy corresponding to their larger districts, this almost doubling of the average amount of committee work for West Virginia senators is extreme. The relatively large number of committee assignments in the Senate works against the purpose of dividing up labor in the Senate in the first place.

In an optimal committee system, not only would members be evenly dispersed across the committee system, but the workload of the committees also would be evenly distributed (Rosenthal 1974a). An even workload allows a legislature to gain the maximum benetit from its division of labor by committee. Each committee would be kept busy and none would be overloaded.

Because of a lack of comparable data, comparisons among state legislature's committee workloads cannot be made. However, West Virginia's state legislature has evolved a practice that works against the even distribution of work among its committees. In both the West Virginia House and Senate there is a tacit designation of primary and secondary substantive committees. The primary committees, Finance, Education, Judiciary, and Government Organization, receive most of the important legislation and are allotted the bulk of the time for their work. In this sense, the chambers each currently have four important substantive committees, instead of 16 (Senate) or 12 (House) of equal status. This system may expedite the consideration of key legislation, but it does not take full advantage of the possible benefits that the division of labor into committees can provide.

Two aspects of this primary/secondary committee arrangement likely mitigate any significant loss of efficiency, however. First, each member is traditionally appointed to at least one of the key committees. This practice provides every member an opportunity to get involved in a primary committee. Also, legislators traditionally serve on no more than one of these primary committees in the House and two in the Senate. This prevents any member from becoming overburdened by these chores. The second mitigating factor is the regular use of subcommittees to further divide up the workload of the important committees. Subcommittees allow members to specialize on their subcommittee work, as is done in the U.S. Congress. The use of subcommittees is an important, positive aspect of the West Virginia state legislative committee system, one that is seen in only 22 other state Senates, and 29 other state Houses (American Society of Legislative Clerks and Secretaries 1988).

The amount of experience committee members and chairs have on their committees is an important resource for state legislatures. In the U.S. Congress, members have an incentive to stay on the same committees throughout their careers, as internal committee power is delegated on the basis of committee seniority. But this is not the case in most state legislatures. Committee chairs and assignments are handed out by the state legislature's leadership on the basis of loyalty to the leadership, expertise, and other reasons not related to committee seniority (Rosenthal 1981). This practice creates greater turnover in state legislature's committees than in the committees in the U.S. Congress. Moreover, state legislators typically serve in their chamber for far shorter periods than their federal counterparts, further reducing the possibility of having stable, experienced committees. This is especially true for West Virginia, which has extremely high legislative turnover rates. During the 1991-92 West Virginia state legislature, of the 26 House and Senate substantive committees in existence for more than one session, only one of the committee chairs had more committee experience than anyone else on the committee (Kopp 1991).

Although West Virginia's extremely high turnover rate does not bode well for committee stability, there is a tradition here of members staying on the same committees throughout their careers, with less committeehopping than is found in most states. For example, from 1974 to 1977 the West Virginia state legislature ranked much lower in committee turnover (38th) than it typically did in overall legislative turnover (Oxendale 1981). While no recent comparative data exist in this regard, in the 1991-92 legislature the turnover in West Virginia's legislative committees was only moderately greater than the overall turnover in the body as a whole (Kopp 1991). In the House, an average of 50 percent of the members of each primary committee and 55 percent of each secondary committee members had no previous experience on that committee. In the Senate, the rates were somewhat lower, reflecting the lower overall turnover in that chamber. Thirty-five percent of the members of the Senate's primary committees and 43 percent of its secondary committees had no previous experience on those committees. This suggests that while a significant number of new committee members come on board with each new legislature, there is usually a solid block of experienced members to provide continuity and expertise.

The final aspect of the West Virginia state legislative committee system to be considered is the level of autonomy committees are accorded in relation to the chamber's leadership. Highly autonomous committees have a firm control over their own agendas and a large degree of input concerning the laws the legislature produces in their policy areas. This gives both committee members and other actors interested in that policy area an incentive to work within the committee structure, where the decisions are made by legislative specialists.

In the U.S. Congress, committees are extremely autonomous, and the chamber leadership's role is mainly one of traffic control. State legislatures, on the other hand, have traditionally been dominated by their chamber leaders (Keefe and Ogul 1989). This is not necessarily bad, however, because a strong leader can coordinate the activities and output of the entire legislature, resulting in more efficient law-making.

In West Virginia, as in almost all states, committee assignments and the designation of chairs are made almost entirely at the discretion of the Speaker of the House and the President of the Senate in their respective chambers. Committee chairs in the West Virginia legislature typically run their committees with a firm hand on the agenda, but they owe their positions to the chamber leadership and are responsive to it.

Also, as in most other states, the referral of bills for consideration by one committee or another is controlled by the chamber's leadership. While many bills are assigned routinely, the lack of clear and established committee jurisdictions allows a leader the opportunity to send a bill to either a friendly or an unfriendly committee. The presiding officers in the West Virginia state legislature, as in many states, are also able to refer legislation to morethan one committee, anothertool allowing them to influence the committee system's output.

Uniform rules for committees that make clear and standard the powers and duties of each committee member are seen as a way to keep a committee from being dominated by its chairperson and (indirectly) by the chamber leadership (Burns 1971; Reeves 1985). In this respect, the West Virginia legislature is at a relative disadvantage when compared to other states. Thirty-six other legislatures have uniform committee rules for at least one chamber, but neither the West Virginia House of Delegates northe Senate has such rules.

The committee system in West Virginia is therefore quite close to the average system around the country. The number of committees is small enough to reduce complexity, and keep the number of members' assignments manageable (at least in the House). Although the workload is spread out unevenly among the committees, certain aspects of the primary/secondary committee arrangement moderate the inefficiencies it causes. And finally, the committees in West Virginia, like the committees in most other state legislatures, are dominated by the chamber leadership.

Legislative Processing

A direct assessment of the legislature's law-making function can be made by examining certain aspects of the overall legislative process itself. For example, legislatures can set up certain rules and mechanisms that expedite and streamline bill processing. The Council of State Governments (CSG) has identified seven mechanisms to accomplish these goals: establishing deadlines for bill introduction, for committee action on bills, and for action on bills in both chambers; allowing the pre-filing of bills and companion bills; allowing the use of committee bills; and allowing bills to carryover from the first to the second session (CSG 1990). These mechanisms make bill processing more efficient, by reducing duplication of effort, regularizing the process, and decreasing end-of-session logjams (Burns 1971; Reeves 1985).

To compare state legislatures in this regard, a scale was developed that rates them on how many of the seven mechanisms each chamber uses. One point is given to a legislature if either its House or Senate uses each one of the seven mechanisms, so that each state could score from zero to 14 points. While the median score in the U.S. is 7.5 on this scale, West Virginia's legislature scores a 10, and is tied for the sixth highest score in the country. In West Virginia, the House of Delegates uses all seven of the mechanisms mentioned above, while Senate rules allow for the pre-f iling of bills, deadlines for bill introduction, and the use of companion bills. So on this aspect of law-making, West Virginia ranks among the top legislatures in the country.

The quantity of legislation introduced in state legislatures has increased dramatically since the 1960s due to increased population density and heterogeneity, New Federalism's devolution of federal responsibilities to the states, legislative professionalization, and other factors (Rosenthal and Forth 1978). Some state legislatures have responded to this increased workload by essentially abandoning their deliberative role, and passing legislation into law at a high rate. These tend to be some of the least prof essionalized state legislatures. For example, the Virginia, South Dakota, North Dakota, Nevada, and Montana legislatures pass into law over 60 percent of the legislation introduced to them. Arkansas' passage rate is an astonishing 78 percent! Clearly, these state legislatures have largely become "legislative assembly lines" where relatively little consideration is given to each of the laws they pass (Rosenthal 1981).

West Virginia, on the other hand, ranks highly in this regard. During the 1989 legislative session, West Virginia's legislature considered 1,699 bills and resolutions, close to the national average, and passed 17.7 percent of them. This was much lower than the national average of 35 percent (CSG 1990). This passage rate was the sixth lowest in the nation, surpassed largely by only the most professional legislatures, such as Massachusetts, New York and New Jersey. If it is assumed that the quality of the bills introduced is reasonably constant across the states, this indicates that the West Virginia legislature takes its deliberative role as seriously as any legislature in the country, working hard to protect the state from poorly crafted or ill-considered proposals.

Another way of looking at the level of consideration each bill receives in the legislature is the number of bills

passed per legislative workday. The more bills that are passed each day, the less consideration each is likely to receive. While the length of the deliberation does not necessarily equal the quality of deliberation, these are likely to be highly correlated. Using the total calendar days of the 1988 and 1989 legislative sessions, the median U.S. legislature produced 4.7 laws per day. West Virginia's production rate of 3.5 bills per day is lower than average, again boding well for the level of legislative deliberation here.

Oversight of the Bureaucracy

The state legislature has a duty to oversee the administration of state law as it is carried out by the executive branch. Until the 1970s, this duty was "only dimly perceived and spasmodically practiced" by most state legislatures (Rosenthal 1981). However, many state legislatures now attempt to oversee the bureaucracy more seriously because they have come to realize its great influence and discretion in implementing state public policy. However, assessing a state legislature's oversight activities is quite difficult because oversight is not routinely practiced and rarely recorded. This article examines those aspects of oversight for which comparative data exist. Specifically, it examines the powers available to those committees that have the responsibility to review administrative rules, certain aspects of the legislature's role in the budget process, and the existence of "sunset legislation" regarding executive agencies.

Executive agencies in states routinely promulgate administrative rules in the process of implementing public policy. These rules have the force of law, and often have a serious impact on a state's citizens and businesses. For example, while the legislature may charge the Division of Natural Resources with the regulation of hunting in the state, it is the Division, through its administrative rules, that determines precisely where hunting will be allowed, the dates of permissible hunting for various types of animals, and bag limits. The legislature's ability to supervise the promulgation of these rules is critical to its ability to control the public policy pursued by the state.

The Council of State Governments has identified five powers that legislative committees charged with overseeing administrative rules can have, including whether it can review existing and/or proposed rules and whether it can suspend a rule with or without the full legislature's agreement. The West Virginia Joint Legislative RuleMaking Review Committee has the primary responsibility for oversight in the state legislature. It has only two of these five powers (the ability to review proposed rules and the requirement of explicitly approving each rule), which is the median level for all state legislatures.

Sunset laws require a thorough, periodic reevaluation of an agency (typically every 5-10 years), with an explicit legislative reauthorization if the agency is to continue. Sunset legislation was designed both to force legislatures to oversee agencies more closely (if only periodically), and to cause agencies to be more aware of their accountability to the legislature (Rosenthal 1981). While the impact of these reviews has been less than the promise, it is still an institutional mechanism that provides an incentive for legislative oversight. In 1990, West Virginia was one of 25 states that had regularized sunset legislation (CSG 1990). However, ten states had comprehensive sunset legislation, while West Virginia's legislation applies only to selective agencies. This places the West Virginia state legislature in the median category nationally.

In almost every state legislature, the consideration of the governor's proposed budget for the state is the session's single most important activity (Rosenthal 1990). While the budget is important for a variety of reasons, it is certainly one of the best tools the legislature has to control what both the governor and the bureaucracy do. Careful consideration of the budget by the legislature increases its ability to oversee the executive branch.

In the West Virginia state legislature, the budget is taken very seriously and given as much attention as is possible in its limited session. One indicator of this is the level of legislative experience of those serving on the House and Senate Finance Committees, which are given principal responsibility for the budget bill. These committees have legislators with greater experience in both the legislature and on these key committees than is the norm for their chambers (Kopp 1991). The average number of years of legislative service was 3.8 for all members of the 1991-92 House of Delegates, whereas those serving on the Finance Committee averaged 5.1 years. In the Senate, the numbers were 7.7 and 8.9 years, respectively. Perhaps even more telling is the fact that only three of 25 members of the 1991-92 House Finance Committee had no state legislative experience at all. This means that only 12 percent of this committee were rookies, as compared to 31 percent of the full House. In the Senate, only one first-year Senator was appointed to Finance, giving that committee 6 percent freshmen versus 15 percent in the full chamber.

The Finance Committees' members also have a higher level of on-committee experience than almost any other legislative committee. At the opening of the 1991 legislative session, House Finance Committee members averaged 2.4 years on that committee, as compared to an average of 1.6 years for all other substantive committees. Only the Judiciary Committee, with an average of 2.8 years of on-committee experience, surpassed Finance in this regard. In the Senate, Finance Committee members averaged 4.8 years of committee experience, while the average for other Senate committees was 2.6 years. Only the Health and Human Resources Committee was more experienced,

with an average of 5.2 years of committee service per member. Members of both Finance Committees therefore had consistently more committee and legislative experience than was the norm in their chambers. This indicates that these committees have greater than average legislative resources to draw upon in conducting their important business.

The Finance Committees are also extremely powerful through their influence on the definition of "legislative intent" regarding the budget. Because the state government is so massive and complex, it is impossible for all bureaucratic discretion to be purged from the budget. But after the budget is passed by the legislature and signed by the governor, the House and Senate Finance Committees meet jointly to write the "Digest of the Enrolled Budget Bill" (Legislative Auditor 1991). This document formally defines in more explicit detail than the budget bill how the legislature intends for agencies to spend their funds. This gives the Finance Committees the power to control bureaucratic spending discretion to the degree the committee desires. While the Supreme Court of Appeals has held that this document does not have the force of law, there is certainly powerful incentive for bureaucrats to consider its directives very carefully (Common Cause v. Tomblin 1991). This process is controversial because of the power it gives to a small group of legislators, but it is undoubtedly a potent way for the Finance Committees to control the bureaucracy's activities.

Another factor influencing the legislature's budget review capacity is the governor's deadline for submitting his/her budget proposal to the legislature for consideration. This is particularly crucial to the many states, like West Virginia, whose legislatures meet for a short and constitutionally limited time. The later the governor is allowed to submit his/her proposal to the legislature, the less time the legislature has to review and make its mark upon it. A 500-plus-page, multibillion-dollar budget bill and supporting documentation can get barely more than a cursory inspection and a legislative rubber stamp in even as much as three or four weeks.

In this respect, the West Virginia legislature is again in the middle of the pack, being one of 15 states that require the governor to submit the budget proposal during the first week of the legislative session (CSG 1990). Ten states give the legislature somewhat more time by requiring submission prior to the convening of the regular legislative session, but 25 states allow the governor even more time than West Virginia does.

The development and distribution of fiscal notes to legislators also influences the quality of the legislature's budget review. These memos are prepared by either executive agencies or legislative staff to describe the various fiscal impacts of proposed legislation. Without this kind of information, legislators may be convinced to go along with a "good idea" without realizing its current and future costs. Fiscal notes provide legislators with the information needed to weigh the costs along with the benefits of any proposed changes in the law or in state programs.

The fiscal notes provided by the West Virginia state legislature for its members have both good and bad points, First, they are among the most comprehensive in the country, providing legislators information on the purpose of each proposal, the costs involved, the projected future costs, the source of the revenue to meet these costs, and (in the House) the fiscal impact on local governments (CSG 1990). Only 13 other state legislatures are provided with all of this information. Further, West Virginia is one of only 27 states that provide these notes to all of its members on all proposals with any fiscal content, whereas 23 states only provide them for a subset of their members and/or on a subset of bills. This means that there is no monopoly on this information in the West Virginia state legislature.

One problem with the fiscal notes used in the West Virginia legislature is that they are prepared not by independent legislative staff, but by the executive agencies who will execute the programs if funded. This is primarily because of a dearth of professional staff in the legislature. Agencies are thereby given an opportunity to affect the outcome of legislative decision-making in a way favorable to their interests. Cost estimation is an inexact science, and the decision to be conservative or liberal in these estimations may hinge upon whether an agency wants to see the proposal passed or not. While agencies' fiscal notes are routinely checked for accuracy and thoroughness, and any breach of faith with the legislature would be devastating to an agency's credibility (and future budget!), this discretion may sometimes be important.

Legislative oversight of the executive branch is difficult to document, especially in ways that allow for cross-state comparisons. However, the fragmentary evidence developed here indicates that the West Virginia state legislature understands its oversight responsibilities and pursues them in a moderately rigorous way. It appears to oversee the bureaucracy at a level comparable to the U.S. average, even despite its belowaverage legislative resources.

Representation

The legislature is the branch of government that is closest to the people, and as such it is the heart of representative democracy in the states. Individual state legislators represent far fewer citizens than the governor and other statewide elected officials. This means that legislators should be the state officials who are best able to assess and to reflect the opinions of their constituents. It also means that citizens have the best opportunity to assess and to influence these officials' behavior. The quality of the legislature's representation of the state's citizens is therefore of primary importance

in evaluating the legislature. The evaluation of how well West Virginia legislators represent their constituents requires the consideration of at least two of these dimensions: how well the legislature reflects the politically relevant characteristics of its constituents, and the number of constituents each representative serves. As will be shown, West Virginia's state legislature performs poorly on this aspect of its responsibilities, in large part due to the extensive use of multimember districts for electing Delegates and Senators.

West Virginia state legislators, like other elected officials in the U.S., are far more likely to be male, white, and from the dominant political party than the average citizen. Both women and racial minorities are underrepresented in comparison with their proportion in the population (National Conference of State Legislatures 1991). In 1991-92, 19.4 percent of West Virginia legislators were women, slightly above the national average of 18.1 percent. African-Americans made up only 1.5 percent of the West Virginia body, compared to 5.8 percent for all U.S. legislatures. While this shows that there were fewer African-Americans in the West Virginia legislature than the typical state legislature, West Virginia's African-American population (2.9 percent) is also far less than the U.S. average (12.4 percent) (U.S. Bureau of the Census 1989). So while the West Virginia legislature is unrepresentative of the state both in terms of gender and race, the average U.S. state legislature underrepresents women and African-Americans at about the same rate as West Virginia's does. Further, women and African-Americans are better represented in the state legislature today than in the past, and they are much better represented than in statewide offices. For example, compared to the 1967-68 West Virginia legislature's 5.2 percent women and .7 percent AfricanAmericans, today's legislature is much more representative of the general population (Blankenship 1967). Also, with the exception of a Supreme Court of Appeals justice, no woman or African-American has ever been elected to a statewide office in West Virginia.

The West Virginia state legislature is also quite similar to other state legislatures in its unrepresentativeness in terms of its members' occupations. Lawyers and business owners are found far more commonly in U.S. state legislatures than in the population generally (Bazar 1987), and this holds true for West Virginia (see Table 1). In 1986, working people ("employees" in Table 1) were greatly underrepresented, with only 7 percent in the West Virginia legislature and 5 percent nationwide. This disparity is understandable given the sporadic demands for time and money required in both running for and serving in the state legislature. Lawyers and business owners often have more flexible schedules and more surplus resources to accommodate these demands than others. Also, service in the legislature can be advantageous to the careers of lawyers and certain types of businesspersons (Barber 1965).

Another characteristic that impacts the West Virginia state legislature's representativeness is political party affiliation. West Virginia is a solidly Democratic state with a sizable minority of Republicans. In 1992, 66.2 percent of the state's voters were registered Democrats, 30.6 percent were Republicans, and 3.2 percent were either independents or affiliated with a minor political party (Secretary of State 1993). In the state legislature, however, Democrats dominate to an even greater degree than in the population. In the election of 1992, for example, voters picked a House of Delegates composed of 79 percent Democrats and a state Senate of 94.1 percent Democrats.

The state legislature is not representative of the general electorate's party affiliation primarily because registered Republican voters are relatively evenly distributed around the state at the county level, and counties tend to be the basis for state legislative districts. For example, Republicans outnumber Democrats in only six small counties (Doddridge, Grant, Morgan, Ritchie, Tyler and Upshur), with the combined voter registration in these counties being only 4.3 percent of the state's total (Secretary of State 1993). In most counties, Democratic and Republican strength more or less roughly reflects the statewide two-to-one split. Since party affiliation is a strong influence on the vote for state legislators, relatively few Republicans get elected.

It is not clear, however, that Republican strength below the county level is so evenly distributed. If pockets of Republican strength do exist, for example in the suburbs of Charleston, Huntington or Morgantown, smaller districts might allow Republicans to capture a proportion of seats more reflective of their statewide strength. This possibility is illustrated by the fact that the state Senate, with its larger districts, typically has a less representative proportion of Republicans than does the House of Delegates.

In addition to reflecting the relevant characteristics of the state's citizens, the quality of legislative representation can also be affected by the number of citizens a legislator must serve. The extensive use of multi-member legislative districts in West Virginia causes problems here. If representation entails knowing the district's needs and preferences, and reflecting them accurately in the

legislature, large districts (in terms of population) are more difficult to represent than small districts. As the number of constituents increases, there are more opinions to understand and to reflect. But perhaps more importantly, the more populous a district is, the more likely it is that it will be heterogeneous in politically relevant ways. For example, under the districting plan enacted in 1992 the 5th West Virginia House of Delegates district in Wetzel County has approximately 18,000 citizens, most of whom were white, Democratic, and long-time, rural and small-town residents. But the 30th House district in Kanawha County has approximately 121,000 citizens, with a much more diverse set of interests and needs. The rural poor, coal miners, chemical workers, urban interests, government workers, and others all compete for the ear of a Kanawha delegate.' Representation is more difficult for the Kanawha delegate both because he/she must work to understand these often conflicting interests and points of view, and because he/she must then choose between them. This means that the more heterogeneous a district is, the more likely it is that certain citizens may not be understood or may be ignored by a legislator in favor of some competing group.

Because state legislatures throughout the nation have approximately the same number of members while varying widely in population, rural state legislatures generally have an advantage in this aspect of representation (Bryan 1981). For example, each California state senator represented about 744,001 citizens in 1990, while state senators in Montana had only 15,981 constituents to represent. In this respect, one would expect West Virginia to do well, considering its small population.

However, the use of multi-member legislative districts in West Virginia negates this potential advantage. Multi-member districts are those in which more than one representative serves. Because of the U.S. constitutional requirement that representation of citizens be equal in a given chamber, any district in which, for example, two members serve must have a population equal to twice that of a district in which only one member serves. For example, in the 5th West Virginia House of Delegates district, only one delegate is elected by the 18,240 citizens of Wetzel County (1990 population). In the 30th district seven delegates serve about seven times as many citizens (120,561 in 1990) in Kanawha County.

In the West Virginia House of Delegates, districts are represented by from one to seven delegates, with an average of 1.8 per district under the recently passed districting plan for the 1990s. The most populous districts are located in the areas of heaviest population concentration, such as in Wood, Kanawha and Cabell counties. Under the 1982 districting plan (the last plan for which complete comparative data is available), the average number of citizens a delegate represents is 80,318. This is far above the national average of 43,677, and is comparable to such large states as Michigan and Illinois. With a single-member districting plan, the average delegate in the 1980s would have represented only about 19,360 people, less than one-half the national average, This would have given the West Virginia House of Delegates one of the best representation ratios in the country.

Multi-member districts also disrupt the link between citizen and legislator in a variety of other ways. The amount of effort required to cast a thoughtful vote in a state legislative election is multiplied by the number of candidates on the ballot. For example, a Wetzel County voter typically needs to evaluate only two major party candidates for the House of Delegates in the general election, whereas the Kanawha County resident is often required to evaluate 14. Given a limited amount of time and energy to devote to casting a vote, Wetzel County residents are likely to be far more informed about the candidates than those of Kanawha County. Poorly informed voting leads to poor representation, increasing the likelihood that those entering office will not reflect the opinions and values of their constituents.

Multi-member districts may also disadvantage certain groups in a state and/or district. For example, as discussed earlier, enclaves of a minority party may be more likely to elect representatives in smaller districts. Also, certain economic or regional minorities may be concentrated in small areas of a large district and therefore be disadvantaged by multi-member districting. For example under the 1982 districting plan, in the 23rd Delegate district (Kanawha County) in 1991-92, all 12 members lived in either Charleston or South Charleston, even though these cities comprised only 34.5 percent of that district's population. In other words, Charleston and South Charleston had 34.5 percent of the population, but gained 100 percent of the representation. In fact, in the 60 House seat races in this district under the 1982 districting plan, only three were won by someone living outside of Charleston or South Charleston. This is not to say that legislators respond only to citizens from their hometown, and certainly there were delegates from the 23rd who cared about the citizens of Pinch, St. Albans and Nitro. But if the dominant areas have interests that are systematically different from these other areas, multi-member districting may lead to these interests being unrepresented. At the very least, the impression of being represented by "one of our own" is important to sustaining faith in government, and this is threatened by multi-member districting.

These and other problems with multi-member districting led the federal government to outlaw them for congressional seats in 1842 and over 60 percent of the state legislatures that used them in 1960 had switched

to single-member districting plans by 1982 (Niemi, Hill and Grofman 1985; Benenson 1990). Following the 1992 round of redistricting, only four other state senates and 10 other state houses have any multi-member districts, and very few of these states use them to the extent to which they are used in West Virginia. In states that switched to single-member districts, not only was the average size of the districts minimized, but interest group and political party influence was reduced, as was the rate at which legislators voluntarily retired from office (Jewell 1982; Niemi and Winsky 1987). All of these effects bode well for representation. And perhaps just as important for the West Virginia state legislature, these benefits could be reaped without a single tax dollar being spent.

There are two primary reasons why the West Virginia state legislature persists in using multi-member districts, one constitutional and one practical (Schochet and Hedge n.d.). The West Virginia state constitution mandates that all state senate districts be bounded by county lines and be equal in population (West Virginia Constitution Art. 6, sec. 4). With population differences between counties, these requirements mean that some counties (or groups of counties in a single senatorial district) will need to elect more than one senator. While the House of Delegates is under no explicit requirement to be districted along county lines, the tradition has been that these boundaries are used as at least the starting point for the districting of that chamber, as well. However, in recent decades, the county line has come to be less inviolable than the equal representation requirement, especially since the U.S. Supreme Court mandated the latter for all U.S. state legislative districts in Gray v. Sanders (1963). In both the 1982 and 1992 redistricting plans, for instance, there were many violations of the county line rule, as bits and pieces of counties were cut off to balance district populations.

The practical reason multi-member districts continue to be used in West Virginia is that it is easier to carve up the state into fewer districts that are proportional in population than two or three times as many districts that are equal in population. However, with the advances in computer-based census information and new redistricting software, this is certainly less of a consideration today than in past decades (Benenson 1990).

The practical obstacles to single-member districting are surmountable. The question of whether or not to continue the tradition of multi-member districting therefore hinges upon a judgment about whether it is more important to maintain the integrity of county boundaries and protect the political fortunes of certain incumbents who may feel threatened by a shift to single member districts, or to address the problems of representation that these multi-member districts cause.

Conclusions

Legislative capacity is a multidimensional and relative concept. By breaking it down into its component parts and using cross-state comparisons, this article has evaluated many important aspects of the capacity of the West Virginia state legislature. In many respects, especially in its committee system and oversight activities, it ranks among the middle range of states. For a state such as West Virginia, which ranks low on population levels and near the bottom in per capita income, to achieve this "average" ranking for its state legislature is quite an accomplishment. And in the area of legislation processing, the West Virginia state legislature ranks as one of the top legislatures in the country.

One area where the West Virginia legislature lags far behind other state legislatures is in the area of lawmaking resources. With little money to spend, and being unable to take advantage of the economies of scale available to more populous states, West Virginia ranks below average on legislative salaries, staff, and overall expenditures. This accounts for the traditionally low ranking of the state's legislature on scales that use these factors as the main criteria for evaluation (Grumm 1971; Squire 1992). On the other hand, the effort brought to bear on providing the legislature with resources, in terms of the level of available resources that are spent on the legislature, is admirably high. Unfortunately, this indicates that there is probably less room for increasing legislative resources in West Virginia than in most states.

The area of representation is both where the West Virginia state legislature does the worst relative to its counterparts in other states, and where there is the greatest potential for improvement. Little can be done directly about increasing the representation of women and African-Americans in the legislature, and West Virginia does no worse than other state legislatures in this aspect of representation. But the use of multi-member legislative districts is an area of both great problem and prospect. These districts not only greatly inflate the average number of citizens a legislator represents, they also disrupt the process of representation in a variety of other ways. The great effort that is brought to bear in providing West Virginia with a competent legislature means little if its members do not represent their constituents well.

The switch to single-member legislative districts would simply and cheaply eliminate these problems. Constitutional roadblocks to this change have likely been sufficiently eroded over the past 20 years, or else they could be lifted through constitutional amendment. Resistance based on tradition would likely be reduced through open debate of the merits of such a change. Political concerns of the majority party being disadvantaged by single-member districts could be allayed by the knowledge that even single-member districting

plans are never politically neutral- in other words, little reason exists for not undertaking this single most important move toward improving the West Virginia state legislature.

The political debate over this issue in the 1992 legislative session was lively, and the current plan moves away from multi-member districts to a moderate degree. For example, the 1982 plan had only 14 single-member delegate districts, while the 1992 plan has 33. Further, the average number of delegates per district was 2.5 under the 1982 plan, but is 1.8 under the 1992 plan. However, West Virginia still uses multi-member districts to a greater extent than almost any other state, and it is unlikely that this will change until the next mandatory round of redistricting, in the year 2002.

The following are other, relatively low-cost changes that could improve the capacity of the West Virginia legislature to fulfill its three major functions:

• Investigate the reasons for the extremely high rate of membership turnover in the West Virginia state legislature, and take steps to encourage members to serve at least a few consecutive terms.

• Develop a more even distribution of the workload among the legislature's committees by eliminating the primary/secondary committee distinction.

• Reduce the number of committee assignments for each state senator to allow for greater focus in committee work.

• Encourage committee membership stability by developing a committee seniority ranking system, perhaps tied to the distribution of such benefits as speaking and voting order, staff and other resource control, and perhaps even committee leadership positions.

• Adopt a uniform set of rules for all the committees in each chamber.

• Require the governor to present his/her budget proposal in advance of the legislative session, allowing for better staff preparation and the use of the entire session for budgetary review.

In the long run, the quality and capacity of the West Virginia state legislature depends upon the quality and commitment of the people elected to it. However, these proposals may help both to continue to attract highquality citizens to the legislature, and to allow them to do their best work for the state once they are there.

References

American Society of Legislative Clerks and Secretaries. 1988. Inside the Legislative Process.
    Denver, CO: National Conference of State Legislatures,

Basehart, Hubert Harry. 1980. "The Effect of Membership Stability on Continuity and Experience in
    U.S. State Legislative Committees." Legislative Studies Quarlerly 5:55-68.

Barber James David. 1965. The Lawmakers: Recruitment and Za'~tation to Legislative Life. New
    Haven, CT: Yale University Press.

Bazar, Beth. 1987. State Legislators' Occupations: A Decade of Change. Denver, CO: National
    Conference of State Legislatures.

Benenson, Robert. 1990. Jigsaw Politics: Shaping the House After the 1990 Census. Washington,
    D.C.: Congressional Quarterly.

Blankenship, C.A. 1967. Ed. Manual of the Senate and House of Delegates: 58th Legislature,
    1967-68. Charleston, WV: West Virginia House of Delegates.

Bowman, Ann O'M,, Richard C. Kearney. 1988. "Dimensions of State Government Capability."
    Western Political Quarterly 41:341-362.

Bryan, Frank M. 1981. Politics in the Rural States: People, Parties and Processes. Boulder, CO:
    Westview.

Burns, John. 1971. The Sometimes Governments. New York: Bantam Books.

Common Cause v. Tomblin. 413 S.E. 2nd 358 (WV 1991).

Council of State Governments (CSG). 1990. The Book of the States, 1990-91 Edition. Vol. 28.
    Lexington, KY: The Council of State Governments.

Francis, Wayne L. 1989. The Legislative Committee Game. Columbus, OH: Ohio State University
    Press.

Gray v, Sanders (372 U.S. 368,1968).

Grumm, John G. 1971. "The Effects of Legislative Structure on Legislative Performance." In State
    and Urban Politics: A Reader in Comparative Public Policy, eds., Richard 1. Hoff erbert and Ira
    Sharkansky. Boston: Little, Brown.

Heard, Alexander, 1966. Ed. State Legislatures in American Politics. Englewood Cliffs, NJ:
    Prentice-Hall.

Herzberg, Donald G., and Alan Rosenthal, 1972. Eds. Strengthening the States: Essays on
    Legislative Reform. New York: Anchor.

Holmes, Darrell E., 1989. Ed. West Virginia Blue Book 1989. Charleston, WV: West Virginia State
    Senate.

Jewell, Malcolm E. 1982. Representation in State Legislatures. Lexington, KY: The University Press
    of Kentucky.

Jewell, Malcolm E., and Penny M. Miller. 1988. The Kentucky Legislature: Two Decades of
    Change. Lexington, KY: The University of Kentucky Press.

Keefe, William J., and Morris S. Ogul. 1989. The American Legislative Process: Congress and the
    States. 7th ed. Englewood Cliffs, NJ: Prentice Hall.

Knight, Erin. 1988. "Legislative Professionalism in the Fifty States: 1967-1984." Paper prepared for
    the annual meetings of the Southwestern Political Science Association, Houston, TX.

Kopp, Donald L. 1991. Ed. Manual of the Senate and House of Delegates: 70th Legislature,
    1991-92. Charleston, WV: West Virginia House of Delegates.

Legislative Auditor. 1991. "Digest of Enrolled Budget Bill: Regular Session of the Legislature of
    West Virginia." Charleston, WV: Legislative Auditor's Office.

LeLoup, Lance T. 1978. "Reassessing the Mediating Impact of Legislative Capability." American
    Political Science Review. 72:616-621.

National Conference of State Legislatures. 1988. "Staff Size in State Legislatures: 1988."
    Photocopy. Denver, CO: National Conference of State Legislatures.

-----. 1991 a. "Black, Hispanic and Women Legislators Following the November 1990 Election."
    Photocopy. Denver, CO: National Conference of State Legislatures.

----- 1991 b. "Review of the Joint Legislative Staff Agencies of the West Virginia Legislature."
    Denver, CO: National Conference of State Legislatures.

Niemi, Richard G., Jeffrey S. Hill, and Bernard Grofman. 1985. "The Impact of Multimember
    Districts on Party Representation in U.S. State Legislatures." Legislative Studies Quarterly
    10:441-455.

Niemi, Richard G., and Laura R. Winsky. 1987. "Membership Turnover in U.S. State Legislatures:
    Trends and Effects of Redistricting." Legislative Studies Quarterly 12:115-124.

Oxendale, James R., Jr. 1981. "Membership Stability on Standing Committees in Legislative Lower
    Chambers." State Government 54:126-129.

Oxendale, James R., and Allan S. Hammock. n.d. "West Virginia: Coal and the New West Virginia
    Politics." Photocopy. Morgantown, WV: West Virginia University.

Reeves, Mavis Mann. 1985. The Question of State Government Capability. Washington, D.C.:
    Advisory Commission on Intergovernmental Relations.

Rosenthal, Alan. 1974a. Legislative Performance in the States. New York: Free Press.

------ 1974b. "Turnover in State Legislatures." American Journal of Political Science 18:609-616.

------ 1981. Legislative Life. New York: Harper and Row.

------ 1989. "The Legislative Institution: Transformed and At-Risk." In The State of the States. Ed.,
    Carl E. Van Horn. Washington, D.C.: Congressional Quarterly. 1990. Governors and
    Legislators: Contending Powers. Washington, D.C.: Congressional Quarterly.

Rosenthal, Alan, and Rod Forth. 1978. "The Legislative Assembly Line: Law Production in the
    American States." Legislative Studies Quarterly 3:265-291.

Schochet, Katrina L., and David M. Hedge. n.d. "Redistricting in West Virginia." Photocopy.
    Morgantown, WV: West Virginia University.

Secretary of State, West Virginia. 1993. "Election 1992 Statistics." Photocopy. Charleston, WV:
    State of West Virginia.

Squire, Peverill. 1988. "Career Opportunities and Membership Stability in Legislatures." Legislative
    Studies Quarterly 13:65-82. 1992. "Legislative Professionalism and Membership Diversity."
    Legislative Studies Quarterly 17:69-79.

U.S. Bureau of the Census. 1989. Statistical Abstract of the United States. 11 Oth ed. Washington,
    D.C.: U.S. Government Printing Office. 1990. Compendium of State Government Finance in
    1989. Washington, D.C.: U.S. Government Printing Off ice.

Van Son, Victoria. 1993. "State Yearbook 1993: Competition by the Numbers." Governing.
    February:35-43.

Weberg, Brian. 1988. "Changes in Legislative Staff." The Journal of State Government. Spring:
    190-197.

Welch, Susan, and Donley T. Studlar. 1990. "Multi-Member Districts and the Representation of
    Women: Evidence from Britain and the United States." Journal of Politics 52:391 412.