State Rainy Day Funds and Fiscal Crises: Rainy Day Funds and the Recession Revisited

Transcription

1 State Rainy Day Funds and Fiscal Crises: Rainy Day Funds and the Recession Revisited JAMES W. DOUGLAS and RONALD KEITH GADDIE The recession of the early 1980s prompted many states to establish budget stabilization (rainy day) funds. Initial examinations of rainy day funds find a limited impact by the funds in alleviating fiscal stress. In this article, we propose an enhanced model of rainy day fund impact. Using data from 48 states for the recession, our analysis indicates that the presence of a number of structural factors and the maintenance of generally large balances in other funds entering recession helps to alleviate fiscal stress when a state s economy is in recession. The recession of the early 1980s imposed a condition of fiscal stress upon the states. Falling revenues forced state officials to cut back on spending and increase taxes in order to fund government services. This prompted many states to establish budget stabilization (rainy day) funds. The number of states with rainy day funds increased from 12 to 38 between 1982 and States creating such funds hoped that savings placed into the funds during good times would enable them to avoid some of the fiscal stress emanating from short-term economic downturns. But do rainy day funds really alleviate fiscal stress? In this article, we examine the existing literature on rainy day fund impact and propose an enhanced model of rainy day fund impact that indicates a significant effect by certain types of funds in alleviating fiscal stress. James W. Douglas is an Assistant Professor in the Department of Government and International Studies, University of South Carolina, Columbia, SC He can be reached by at jdouglas@gwm.sc.edu. Ronald Keith Gaddie is an Associate Professor in the Department of Political Science, University of Oklahoma, 455 West Lindsey Street, Norman, OK He can be reached by at rkgaddi@ou.edu. 1. Russell S. Sobel and Randall G. Holcombe, The Impact of State Rainy Day Funds in Easing State Fiscal Crises during the Recession, Public Budgeting & Finance 16, no. 3 (Fall 1996): Douglas & Gaddie / Rainy Day Funds and Fiscal Crises 19

2 LITERATURE Rainy day funds are potentially useful to state governments because states are compelled to maintain balanced budgets. Most states have either a constitutional or statutory requirement that their budgets remain balanced throughout the fiscal year. Strong incentives to balance budgets, such as political culture and the desire for good credit ratings, exist for states not bound by strict balance requirements. 2 This need to maintain balance in the budget forces states to resort to tax increases and cuts in expenditures during times of fiscal stress, actions that can actually work to worsen the economic situation. 3 Having protected monetary reserves on hand to deal with revenue shortfalls reduces the pressure on elected officials to increase taxes and cut spending and protects taxpayers from revenue ratchet, which occurs when taxes are increased to help a state cope with fiscal stress during a recession but not subsequently decreased when the recession is over. This phenomenon both increases taxes at a time of economic contraction and also places a larger tax burden on taxpayers over the long run. 4 Finally, rainy day funds can be used to help states deal with poor revenue projections and unforeseen emergencies such as flood or tornado damage. Despite their potential importance, little empirical evidence exists regarding the impact rainy day funds have on fiscal stress during periods of recession. The recession of the early 1990s provided an opportunity to assess how rainy day funds perform at reducing fiscal stress. Steven D. Gold assembled a number of case studies analyzing how various states coped with this economic downturn. He found rainy day funds did help some states weather the economic storm, but he also found that most states lacked the political will to make the funds large enough to provide much of a budgetary cushion, 5 indicating that the real effectiveness of rainy day funds resided in their liquidity rather than just their presence. Only one study to date has examined all states to ascertain the ability of rainy day funds to reduce fiscal stress during the recession of the early 1990s. In their 1996 study, Russell S. Sobel and Randall G. Holcombe conclude that the ability of rainy day funds to alleviate fiscal stress depends on how the funds are structured. Their regression analysis of the continental United States revealed that rainy day funds alone did not reduce fiscal stress, but that instead, rainy day funds requiring money to be saved during good economic times alleviated fiscal stress during recession. Sobel and Holcombe also examined the impact of rainy day fund savings caps and withdrawal requirements upon 2. Richard Briffault, Balancing Acts: The Reality behind State Balanced Budget Requirements (New York: Twentieth Century Fund Press, 1996). 3. Irene S. Rubin, The Politics of Public Budgeting: Getting and Spending, Borrowing and Balancing (Chatham, NJ: Chatham House, 1990). 4. John C. Navin and Leo J. Navin, The Optimal Size of Countercyclical Budget Stabilization Funds: The Case Study of Ohio, Public Budgeting & Finance 17, no. 2 (Summer 1997): Steven D. Gold, ed., The Fiscal Crisis of the States (Washington, DC: Georgetown University Press, 1995), Public Budgeting & Finance / Spring 2002

3 fiscal stress but found no statistically significant relationship to exist between those requirements and the effectiveness of the rainy day fund. 6 This early model, although a significant starting point, had weak explained variance, leaving room for further specification. 7 Further, the authors made the assumption that there were structural and legal features that influenced rainy day fund effectiveness, implicitly because those features insured the deposition, maintenance, and protection of rainy day fund balances. Yet they never tested the relationship between rainy day fund balances and the relief of fiscal stress, either alone or in competition with the structural and legal variables that are the foundation of their model. In this article we revise Sobel and Holcombe s model by modifying the withdrawal requirement variable, introducing several additional variables, and eliminating the savings cap variables. We also introduce controls for rainy day fund balances and other surpluses held by the state. Our revisions improve upon the findings of the original model and result in a substantial improvement in reducing predictive error over the original model. MODEL SPECIFICATIONS Assessing the impact of rainy day funds is limited by data. During the recession during Ronald Reagan s presidency, only 12 states had rainy day funds, and although 26 other states created contingency funds during the 1980s, there has only been one subsequent recession at the time of this writing. We are therefore constrained, like Sobel and Holcombe, to examining a cross-section of states during one period of time. Dependent Variable To remain consistent with Sobel and Holcombe, we assume a state fiscal policy of neutrality over the business cycle. Sobel and Holcombe state: Under a policy of neutrality, states do not change their tax structures over the business cycle, so revenues move procyclically, and state expenditures grow at a constant rate through the business cycle. If states are constrained against deficit finance (as most are), a policy of neutrality would require building up surplus funds during the good years and using those funds during recessions. States desire rainy day funds to mitigate against the impacts of recessions that require states to reduce their expenditure growth or raise taxes, so the benchmark of neutrality makes sense as a measure of the effectiveness of rainy day funds. 8 By making this assumption, we accept Sobel and Holcombe s operational definition of the dependent variable, FISCAL STRESS. 9 Their operational definition of the dependent 6. Sobel and Holcombe, We would like to thank Russell S. Sobel and Randall G. Holcombe for allowing us access to their data. 8. Sobel and Holcombe, Data for the FISCAL STRESS variable were taken from Sobel and Holcombe. Douglas & Gaddie / Rainy Day Funds and Fiscal Crises 21

4 variable is the sum of tax increases and expenditure shortfall occurring during the recessionary period as a percentage of general expenditures for fiscal year 1988 (the base year prior to the recession). The recessionary period is identified as fiscal years Tax increases are defined as total dollars collected as a result of any legislated tax increases occurring during the recessionary period. Expenditure shortfall is defined as the difference between actual expenditures and the real trend growth rate. 10 If a state s expenditures fell bellow the trend line for a given year, it is considered to have had a shortfall. The shortfalls for each year ( ) are summed to get the total expenditure shortfall for the state. Sobel and Holcombe sum total tax increases and total expenditure shortfall for each state to get the total fiscal stress for each state. Once again, the dependent variable is the total fiscal stress as a percentage of fiscal year 1988 general expenditures for each state. Independent Variables We employ several independent variables in our model. 11 Because we are attempting to determine the ability of rainy day funds to reduce fiscal stress, a dummy variable (RDF) indicating the existence of a rainy day fund for fiscal year 1989 is included. Also included is a measure of the relative size of rainy day funds for each state, RDF%, defined as the total dollars in a state s rainy day fund at the end of fiscal year 1988 as a percentage of fiscal year 1988 general expenditures. This variable is included because it is assumed that having a rainy day fund by itself is insufficient to reduce fiscal stress. Rainy day funds vary in size, and we expect that better-financed funds should perform better at reducing fiscal stress. Including both a dummy variable to control for the presence of the rainy day fund and the RDF% is necessary to more properly capture the effect of the size of the rainy day fund. If we measured the size of contingency funds without this control, the estimation of the regression would treat cases in which there is no rainy day fund (where RDF% necessarily equals 0) like those cases in which a rainy day fund existed but was unfunded. This might lead to temporized slope coefficients and an inaccurate assessment of the impact of the solvency of the fund. The dummy variable controls for the presence of a fund and allows us to obtain a more true reckoning of the impact of RDF%. 12 A number of structural variables are included in the model as well. Certain states specify formulas that are used to ensure that money is placed into their rainy day funds during nonrecessionary periods. Such requirements guarantee that a certain level of funding will 10. Sobel and Holcombe use the period to calculate the real growth trend rate. 11. Data used to construct the rainy day fund, rainy day fund structural provisions, and fund balance variables were taken from selected issues of the National Association of Budget Officers, Budget Processes in the States (Washington, DC, 1987, 1988, 1989) and the National Association of Budget Officers, Fiscal Survey of the States (Washington, DC, 1987, 1988, 1989). We would like to thank Paul Korfonta and Sandra Cannedy of the National Association of State Budget Officers (NASBO) for helping us collect the data for this project. 12. John Chubb, Institutions, the Economy, and the Dynamics of State Elections, American Political Science Review 82, no. 1 (March 1988): Public Budgeting & Finance / Spring 2002

5 be available in rainy day funds when downturns occur in the economy, rather than simply relying on elected officials to allocate resources to the funds on a year-to-year basis. A dummy variable, REQUIRED SAVINGS, is used to identify states that had savings requirements for their rainy day funds prior to fiscal year We expect that REQUIRED SAVINGS will negatively affect fiscal stress. Withdrawal requirements constitute another factor that can have an impact on a rainy day fund s ability to decrease fiscal stress. Several states use formulas to determine when money can be withdrawn from their rainy day funds. These formulas are related to some indicator of fiscal stress, such as revenues not meeting expenditures, a revenue shortfall, a high unemployment rate, or a low growth in adjusted personal income. Their purpose is to ensure that balances in the funds are used to alleviate fiscal stress rather than to fund special projects during good economic times. We use a dummy variable, FORMULA, to denote whether such a formula requirement is used by a particular state. Another form of withdrawal requirement employed in some states is a supermajority vote of the legislature. This type of requirement is also designed to make it difficult for legislatures to withdraw money from their rainy day funds to pay for special projects. Legislatures bound by this constraint, however, are not forbidden to use rainy day fund balances for special projects; the requirement only makes doing so more difficult than it otherwise would be. We use a dummy variable, SUPERMAJORITY, to specify when a state is restricted by a supermajority vote requirement. We expect that both FORMULA and SUPER- MAJORITY will have negative relationships with FISCAL STRESS because states using such requirements should be better able to protect their rainy day fund balances against being raided during good economic times. 13 A final structural variable included in our model is a dummy variable (MULTIPLE FUNDS) indicating whether a state has multiple rainy day funds. 14 Having more than one fund could induce elected officials to set aside more resources for their reserves so that each fund has the appearance of being adequately financed. Additionally, multiple funds may be created for separate purposes. For example, in fiscal year 1988 New Mexico had an operating reserve and a separate school fund reserve. Much like withdrawal requirements, multiple rainy day funds with separate purposes may help protect rainy day fund balances by making it difficult to remove resources from these funds in order to finance special projects that lie outside of the funds jurisdiction. Finally, states with multiple rainy day funds tend to have different withdrawal requirements for the different funds, with one fund being easier to withdraw from than the other. As a result, elected officials might choose to focus any fund raiding upon the rainy day fund with the less-stringent 13. Sobel and Holcombe created an omnibus measure, WITHDRAWAL REQUIREMENTS, that incorporated all of these restrictions on the appropriation of rainy day funds. 14. Oklahoma was scored as having multiple rainy day funds even though it has only one such fund. Oklahoma s rainy day fund operates as two funds in reality. Half of the fund is designated to deal with revenue shortfalls, whereas the other half is earmarked for emergencies. Withdrawal requirements are different for each half of the fund. Douglas & Gaddie / Rainy Day Funds and Fiscal Crises 23

6 withdrawal requirement, leaving the balance in the other fund alone. This would ensure that at least part of the reserve balance would be protected from being plundered during good economic times. Therefore, we expect that MULTIPLE FUNDS will have a negative effect upon FISCAL STRESS. Two control variables are also included in the model. The first measures the depth of the recession in each state. Obviously, some states were affected more by the recession than others. If the severity of the recession upon the states is not controlled for, then it is difficult to determine whether low fiscal stress is the result of the rainy day fund variables or a mild recessionary effect. Once again, to remain consistent with Sobel and Holcombe we accept their operational definition of recession severity. They defined the variable SEVERITY as [t]he rate of growth in state tax revenues during the recession, subtracting out any funds derived from discretionary tax increases. 15 They reversed the sign so that higher values of SEVERITY would be positively related to FISCAL STRESS. Sobel and Holcombe s definition assumes that states with slower revenue growth are more affected by the recession than states with faster growth. The final variable in the model is a control for balances in state accounts other than the rainy day funds. Most states maintain positive balances in other funds, especially the general fund, to help protect themselves against potential revenue shortfalls. The problem with such balances is that they are easily accessible to elected officials. Therefore, maintaining balances in such funds sufficient to ward off the fiscal stress associated with a recession is politically difficult. Some states do keep substantial balances in these funds, and they can be used to reduce fiscal stress during economic downturns. Ignoring these balances would make it difficult to determine whether low fiscal stress is the result of the rainy day fund variables or large reserves in other funds; therefore a control for other funds is warranted. To alleviate this problem we constructed the variable OTHER BALANCES, defined as the sum of all balances for a state except rainy day fund balances as a percentage of the state s fiscal year 1988 expenditures. We expect OTHER BALANCES to have a negative impact upon FISCAL STRESS. Three variables used by Sobel and Holcombe (withdrawal requirement, cap on rainy day fund balance, and rainy day fund cap as a percentage of the budget) were dropped from our model. Sobel and Holcombe used a withdrawal requirement variable to indicate which states were restricted from using regular legislative action to take money out of their rainy day funds. As stated above, they did not find this variable to have an impact upon fiscal stress. We divided this variable into two separate variables, FORMULA and SUPERMAJORITY. This was done because formula and supermajority requirements are different in nature and, therefore, could have different impacts upon FISCAL STRESS. MULTIPLE FUNDS also partially captures the effect of having a withdrawal requirement, because most states using more than one rainy day fund have different withdrawal criteria for each fund. 15. Sobel and Holcombe, Public Budgeting & Finance / Spring 2002

7 Sobel and Holcombe also examined the influence of caps placed upon rainy day fund balances. They used a dummy variable indicating the existence of a cap and an interval level variable denoting the size of the cap as a percentage of the budget (CAP%). Neither of these variables proved to be significant in their model. We chose not to include either variable in our model. We believe that it is more important to measure the actual size of the rainy day fund than the existence or size of a cap. Therefore, we have substituted the cap variables with RDF%. This makes sense given that the existence of a cap does not mean that a state sets the cap as a target. In fact, no state in this study was actually at its cap level. ANALYSIS AND DISCUSSION Descriptive statistics of our variables appear in Table The average fiscal stress confronted by the states examined (25.83) is substantial. Of the 48 states examined, over three quarters (37 states) had rainy day funds at the beginning of fiscal year However, only about 40 percent of cases (19 states) had a formula restriction on withdrawing funds, and less than 10 percent (four states) had a supermajority requirement for withdrawing funds from the RDF. Just over 10 percent of the cases (five states) actually had multiple rainy day funds, and only a quarter (12 states) had rainy day funds and also required the legislature to deposit money into the fund. The average size of rainy day funds, relative to general expenditures, was small (just over one percent). In Table 2 we present the results of Sobel and Holcombe s model of rainy day fund impact on fiscal stress. 17 Once again, their findings indicate that it is not rainy day funds per se that help alleviate fiscal stress. Rather, it is the legal requirement for putting savings into the rainy day fund that alleviates fiscal stress. The only other significant predictor of the level of fiscal stress in the states is the severity of recession in the state, with more severe recession inducing greater fiscal stress. None of the other variables (the rainy day fund dummy variable, the cap on RDF balance, the cap as a percentage of balance, or the existence of a constraint on the legislature to withdraw funds) attain statistical significance. The model yields an R 2 of We present our revised model of rainy day fund impact on fiscal stress in Table 3. As stated above, in addition to the variables retained from Sobel and Holcombe s 16. Alaska and Hawaii are excluded from the analysis. 17. Because we used a different data source than Sobel and Holcombe for several variables, a small number of coding differences exist between our data and theirs. We also dropped Alaska and Hawaii from the analysis. Alaska was dropped because it maintains a large oil reserve balance that cannot be used for general expenditures, but it would have been included in the OTHER BALANCES variable. Hawaii was dropped because it was insulated largely from the recession. It also appeared that these states were not included in Sobel and Holcombe s original analysis, but this is unclear. 18. Sobol and Holcombe indicated to us through personal communication that the explained variance indicated in Table 2 of their Public Budgeting & Finance article (39) was erroneously reported as R 2 =.52 when in fact their analysis revealed an R 2 =.25 (we report the corrected analysis in our Table 1). Douglas & Gaddie / Rainy Day Funds and Fiscal Crises 25

8 TABLE 1 Descriptive Statistics Percentage of Cases Mean SD RAINY DAY FUND FORMULA SUPERMAJORITY 8.33 MULTIPLE FUNDS REQUIRED SAVINGS FISCAL STRESS RDF% SEVERITY OTHER BALANCES model the rainy day fund dummy variable, required savings, and severity of recession in the state we incorporate additional variables to more fully specify the structure of the rainy day funds in the various states. 19 The results of the revised model indicate that the structure of rainy day funds played a substantial role in reducing fiscal stress in states during the recession. The model is also a substantial improvement over the previous model in terms of explanatory power, explaining over 40 percent of the variance in the dependent variable (Adjusted R 2 =.415; R 2 =.514), compared to an R 2 of.25 for the original model. As in Sobel and Holcombe s analysis, neither the rainy day fund nor the controls for withdrawal requirements (which we respecified) were significant in alleviating fiscal stress. The severity of recession in the state (SEVERITY) and the presence of a savings requirement (REQUIRED SAVINGS) significantly affected fiscal stress, as in the previous model. 20 The new variables that most clearly influenced the level of fiscal stress in a state were variables that indicate a general level of solubility, OTHER BALANCES and MULTIPLE FUNDS. The larger the total fund balances in 1992 as a percentage of 1988 expenditures (OTHER BALANCES), the less severe the fiscal stress in the state. This is no great surprise, given that states with larger fund balances should be better able to deal with fiscal stress. What we find to be particularly important is the presence of multiple rainy day funds in the state (MULTIPLE FUNDS). Having more than one rainy day fund 19. An examination of tolerances and variance inflation factors revealed no mulitcollinearity problems. 20. We also re-estimated the regression equation just for the 37 states that had rainy day funds in That analysis (which eliminated the RDF dummy variable) produced nearly identical slope coefficients for all predictor variables and changed neither the direction nor the significance of any independent variable in the model. 26 Public Budgeting & Finance / Spring 2002

9 TABLE 2 Sobel and Holcombe Model of Rainy Day Fund Impact on Fiscal Stress b t Intercept RAINY DAY FUND CAP% CAP REQUIRED SAVINGS SEVERITY WITHDRAWAL REQUIREMENT a R 2 =.25 N = 48 Source: Russell S. Sobel and Randall G. Holcombe, The Impact of State Rainy Day Funds in Easing State Fiscal Crises during the Recession, Public Budgeting & Finance 16, no. 3 (Fall 1996): Notes: Dependent variable: fiscal stress as percentage of 1988 budget. t-statistics in italics are significant at.05 level or better. a Withdrawal requirements include institutional rules that must be followed to withdraw funds from the RDF (outside of a normal legislative act) and formulas that determine if funds can be withdrawn from the RDF. TABLE 3 Revised Model of Rainy Day Fund Impact on Fiscal Stress b SE t Intercept RAINY DAY FUND FORMULA SUPERMAJORITY MULTIPLE FUNDS REQUIRED SAVINGS OTHER BALANCES RDF% SEVERITY R 2 =.514 Adjusted R 2 =.415 N =48 Notes: Dependent variable: fiscal stress as percentage of 1988 budget. t-statistics in italics are significant at.05 level or better. Douglas & Gaddie / Rainy Day Funds and Fiscal Crises 27

10 appears to have had a negative impact upon fiscal stress. States with multiple rainy day funds experienced a reduction in their fiscal stress of 17.5 percent of their fiscal year 1988 expenditures. Thus, it would seem that having multiple rainy day funds can have a strong influence on a state s ability to cope with economic downturns. An initially confounding result in our analysis was the positive relationship between RDF% and fiscal stress. We thought this coefficient might be some statistical by-product of our model specification. Because we include both the proportional size of the rainy day fund balance and the structural and legal factors that theoretically affect the size and security of that balance, it was possible that the structural and legal factors might not attain significance or that some factors might attain significance but result in an insignificant or counterintuitive slope coefficient. The analysis in Table 3 seems to indicate that this is the case, as RDF% is positively associated with the fiscal stress in a state. To attribute this result to model specification is problematic for two reasons: First, the size of the rainy day fund (RDF%) is only weakly associated with the legal and structural variables advanced by Sobel and Holcombe. 21 After further consideration it appears that, at least for the period, rainy day fund balances per se did not significantly mitigate against fiscal stress. We speculate that states where the rainy day fund constitutes a greater percentage of the general expenditure likely created and funded rainy day funds because they were highly susceptible to fiscal stress during recession. Second, states under severe fiscal stress are more likely to cut expenditures in order to balance budgets, meaning that the ratio of their rainy day fund to general expenditures will be more pronounced. So the cruel paradox that states must raise taxes and cut expenditures in recession seems to lead to a paradoxical coefficient: states seemingly better prepared for fiscal shortfalls are most likely to suffer those shortfalls. This explanation is supported by the finding of Philip G. Joyce that in 1997 several of the states with the most volatile budget environments also maintained relatively high balances in their rainy day funds. 22 Why doesn t this explanation hold true for OTHER FUNDS as well? Balances in other funds tend to be much larger (4.7 percent of fiscal year 1988 expenditures) than rainy day funds (1.0 percent of fiscal year 1988 expenditures). It is possible that states with more volatile budget environments are unable to maintain such large balances in their other funds but are able to deposit some resources into their rainy day funds. If budgets are tight, political pressures likely make it difficult to keep large reserves available in more general funds. In Figure 1 we plot the extent of fiscal stress experienced by a state against RDF% and against OTHER FUNDS. Of the 24 states that were above average on RDF%, 11 were also above the average for FISCAL STRESS. The patterns presented in 21. Regressing RDF% onto the variables FORMULA, MULTIFUND, REQUIRED SAVINGS, SUPERMAJORITY, and the RDF dummy variable explained just 16 percent of the variation in RDF%, and the only significant predictors were the RDF dummy variable and the REQUIRED SAVINGS. 22. Philip G. Joyce, What s So Magical about 5 Percent? A Nationwide Look at the Optimal Size of State Rainy Day Funds, Typescript (Syracuse, NY:Syracuse University, n.d.). 28 Public Budgeting & Finance / Spring 2002

11 FIGURE 1 the scatter plot also reveal that, generally speaking, the relationship between OTHER BALANCES and FISCAL STRESS was negative. Of the 22 cases above the average for OTHER BALANCES, only six were also above average for FISCAL STRESS. 23 In sum, the impact of rainy day funds per se on the ability of states to reduce fiscal stress was not as great as we might have assumed. Instead, it seems that the structure of particular funds the inclusion of savings requirements and the existence of multiple funds were related to a greater ability to mediate fiscal stress, even when the presence of a rainy day fund, the solvency of the fund, and the severity of economic recession are controlled for. Of greater importance than the specific rainy day fund savings were the existence of substantial balances outside the rainy day fund that might be used to mitigate against shortfalls. These balances were typically far larger in most states than the rainy day fund balances and therefore provided a more ready source for fiscal relief. CONCLUSIONS A strategy adopted by many states in order to help them cope with short-term periods of fiscal stress is the creation and maintenance of rainy day funds. States hope that rainy day 23. A bivariate regression of FISCAL STRESS onto RDF% revealed an unstable slope coefficient and no explained variance in FISCAL STRESS. A similar regression of FISCAL STRESS onto OTHER BALANCES revealed a negative and significant coefficient and explained about 11 percent of the variation in FISCAL STRESS. Douglas & Gaddie / Rainy Day Funds and Fiscal Crises 29

12 funds will enable them during economic downturns to minimize cuts in service levels and/or increases in taxes until better economic times return. Taxpayers and opponents of big government could also argue that rainy day funds are a tool with which to reduce revenue ratchet. Given the important fiscal role that rainy day funds are expected to play, it is necessary to assess the ability of such funds to alleviate fiscal stress during recessionary times. Previous research by Sobel and Holcombe found that the structure of rainy day funds plays an important role in determining the extent to which the funds are able to reduce fiscal stress. This article has added to this research in a meaningful way. Implicit in Sobel and Holcombe s model is the assumption that structural features of a contingency fund that are associated with the reduction of fiscal stress are therefore indicators of a more effective rainy day fund. These factors savings requirements and withdrawal requirements should promote the creation of substantial and well-endowed funds that can be tapped to make up revenue shortfalls. Therefore, they all imply a larger and better funded rainy day fund relative to the existing budget. When we test the impact of these structural features while also controlling for the solvency of the contingency fund, we discover (1) that most of the variation in the solvency of the fund is not explained by these factors and (2) that factors like multiple contingency funds and savings requirements still exert independent and significant influence in ameliorating the fiscal stress of states when the balance in the contingency fund(s) is controlled. Therefore, we can conclude, at least in the context of the recession, that although the structural features associated with fiscal responsibility in designing rainy day funds reduce fiscal stress, this influence occurs independent of the presumed product of these features, namely, large balances in contingency funds. 30 Public Budgeting & Finance / Spring 2002