Fiscal Problems and Education Finance

Transcription

1 Fiscal Problems and Education Finance by James Alm, Robert D. Buschman, and David L. Sjoquist James Alm is chair of the Department of Economics and Fiscal Research Center at the Andrew Young School of Policy Studies at Georgia State University. Robert D. Buschman is a graduate research assistant in the department. David L. Sjoquist is a professor of economics and director of the Domestic Studies program at the center. This report is the revision of a paper presented at State and Local Finances After the Storm, Is Smooth Sailing Ahead? a conference sponsored by the Tax Policy Center of the Urban Institute and Brookings Institution, held March 30, 2007, in Washington. The report by Nathan B. Anderson on p. 655 is from the same conference. I. Introduction Relative to previous recessions, the 2001 recession was short and not very deep. Even so, it had a significant effect on the fiscal condition of state and local governments. 1 Figure 1 (next page) shows the pattern of own-source revenue in real terms for state and for local governments over the period As can be seen, own-source revenue increased both for state and for local governments until the recession began in mid-2001, at which point state revenue fell dramatically, by 3.4 percent. Although local government own-source revenue did not fall, it did not grow as fast; between 2001 and 2002, local real own-source revenue increased by 1.6 percent, compared with an average of 2.7 percent for the previous 10 years. How did the recession affect state and local government spending on K-12 education in the three years since the recession? That is the issue we examine in this report. Holahan et al. (2004) and others have argued that state and local governments do not like to cut spending on education and that at the time of the recession there was strong public support for states to increase spending on K-12 education. However, during the recent recession, many state and local governments did cut education spending (Holahan 2004). Reschovsky (2004) and Kalambokidis and Reschovsky (2006) considered the effect of the more recent budget shortfalls on education spending. In particular, Reschovsky (2004) documented the change in state fiscal assistance to local school districts in 2003 and He also estimated a current-services education budget so that the reduction in assistance can be compared not just with the previous year s but with what would be required to maintain the same level of education service. Based on the current-services budget, Reschovsky (2004) estimated that for the entire nation, real state aid for education fell by 1.6 percent from fiscal 2003 to fiscal 2004, and by 3.6 percent from fiscal 2002 to fiscal Although real state revenue declined, property tax revenue continued to increase, part of which was presumably used to finance education. (In related work, Ladd (1996) investigated how local districts in Texas and New York responded to the fiscal pressures generated by the economic conditions of the early 1990s; she estimated how the level of fiscal stress affected various budget categories within education.) The remainder of the report is organized as follows. In Section II we track the historic pattern at the national level of state and local spending on K-12 education. We then turn in Section III (p. 640) to a state-level analysis of how K-12 spending was affected by the recession, including correlations between local education spending patterns and state patterns. We conclude in Section IV (p. 647). 2 II. Trends in K-12 Education Expenditures We focus on state and local spending on K-12 education, excluding federal funds to state and local governments for education spending. Because we 1 For a discussion of the effects of the 2001 recession on state fiscal conditions, see McNichol and Harris (2004) and Johnson, Schiess, and Llobrera (2003). Holahan et al. (2004) argue that states faced the most serious fiscal crises since World War II. 2 We do not explore in detail here possible explanations for the observed patterns. In related work, we use simple bivariate regression analysis between possible explanatory variables (e.g., economic, institutional, demographic, political) and changes in spending; we also conduct multivariate regressions to explain these causal factors. See Alm, Buschman, and Sjoquist (2007). State Tax Notes, May 28,

2 Billions $ Figure 1. State and Local Own-Source Revenue (2004 dollars) State are interested in spending by state and by local governments, we measure total spending by total revenue, rather than by state and local expenditures, because it is not possible to separate expenditures by source. In most of our estimates, we also measure spending on a per-student basis, using as our measure of students the fall membership. 3 Those data were obtained from the National Center for Education Statistics (NCES). For years before 2004, data were obtained from the Digest of Education Statistics, available at programs/digest/. For 2004, data were obtained from Overview of Public Elementary and Secondary Students, Staff, Schools, School Districts, Revenues, and Expenditures: School Year and Fiscal Year 2004, available at overview04/. All values are expressed in real (2000 dollars) terms. 4 Figures 2 to 4 (pages ) show U.S. total and per-student real revenue for K-12 education for fiscal 1988 to fiscal Figure 2 presents state plus local 3 For fiscal 2004, fall membership is the only measure of enrollment available, so for consistency, we used it for all years. 4 We use the annual national income and product accounts (NIPA) price index for state and local government to calculate real values. Local government total revenue and total revenue per student, Figure 3 presents only state government data, and Figure 4 presents only local government data. Table 1 shows the annual percentage changes for all measures (for example, total versus per student, total state and local versus state only versus local government only). It should be noted that there was a national recession in and another one the focus of our work here in As can be seen from Figure 2 and Table 1, state plus local government total spending slowed in real terms during the recession. During the rest of the 1990s, real state plus local K-12 revenue increased at generally increasing rates. However, for the last three years of the period, or following the 2001 recession, state and local revenue increased at a much slower rate. Total state and local spending on K-12 education increased by 2.8 percent, and state spending on K-12 education actually declined between 2002 and 2004 by 3 percent; in contrast, local spending increased by 9.9 percent, including a very large increase (9.1 percent) in A similar pattern is seen in per-student spending. During the later part of the 1990s, state spending per student increased more rapidly than did local spending per student, although, as noted above, for local spending increased faster than state spending. The decline in state spending (total and per student) on 638 State Tax Notes, May 28, 2007

3 Table 1. Percent Change in Real Education Revenue (percent change from previous year) Total Revenue Total Revenue per Student Year State plus Local State Local State plus Local State Local % 4.9% 13.8% 8.6% 4.4% 13.3% % 3.0% -0.5% 0.4% 2.1% -1.4% % 4.1% 3.9% 2.3% 2.4% 2.2% % 1.2% 3.8% 0.6% -0.7% 1.9% % 1.8% 3.4% 0.6% -0.2% 1.4% % 0.9% 3.8% 0.8% -0.6% 2.2% % 5.9% -0.9% 1.0% 4.4% -2.3% % 4.7% 1.8% 1.6% 3.0% 0.1% % 5.1% 3.3% 2.5% 3.4% 1.5% % 5.9% 3.6% 3.6% 4.7% 2.4% % 4.3% 2.1% 2.4% 3.4% 1.2% % 4.3% 0.6% 1.9% 3.6% 0.0% % 4.9% 4.3% 3.8% 4.1% 3.5% % 1.2% 1.5% 0.3% 0.2% 0.5% % -0.3% 0.7% -0.9% -1.4% -0.4% % -2.7% 9.1% 1.9% -3.4% 8.3% Source: National Center for Education Statistics. K-12 education is consistent with the observations of Holahan et al. (2004) that spending cuts for fiscal 2004 were more severe than in The recession resulted in a substantial slowing of spending on K-12 education in 2003 and Although the magnitude of changes in education spending during the recession relative to spending in 2000 in figures 2 to 4 is of interest, of perhaps more relevance is the change in real education spending post-2001 as a deviation from trend because that change is more likely to indicate the ways in which the recession affected spending. To calculate those deviations, using NCES data, we estimated for each state a log-linear time trend of real spending per student for each state over the period 1991 through 2001; for the six states that made significant changes in their education funding system during the 1990s, changes that resulted in large shifts in the state share of education spending, we included a dummy variable to reflect the pre- and post-reform years. 6 We then used the trend to estimate the subsequent deviations of real spending from this trend for the years 2002, 2003, and 2004 at the state plus local government level, at the state government level, and at the local government level. Table 2 (p. 642) shows, on a national level, the resulting magnitude of the deviation from trend for state and local spending per student and also for state and for local total spending; the disaggregated results are discussed in Section III. In 2002 real total state and local K-12 spending was slightly above trend. In 2003 total state and local spending fell to 2.2 percent below trend; spending by states and by local school districts also fell below trend, although states were further below trend than were local governments. In 2004 actual total state and local spending increased, so that the percentage below trend in 2004 was slightly less than for However, for 2004 state K-12 spending was even further below trend while local spending was well above trend. A similar pattern is seen in total revenue per student. What those data suggest is that the recession resulted in a substantial slowing of spending on K-12 education in 2003 and Overall, total 5 The National Education Association (2006) provides estimates of (real) total and per-student spending through fiscal The pattern of the association s data generally fits that observed using NCES data. 6 These states were Michigan, Montana, New Hampshire, Oregon, Vermont, and Wisconsin. The detailed results are available upon request. State Tax Notes, May 28,

4 Total, Billions $ Figure 2. State and Local Education Revenue (50 states and D.C., Real 2000 dollars, with trend) Legend: Expenditures per student Total expenditures on K-12 education (in billions) Trend lines Recession periods 1994 state and local and state spending fell from trend between 2002 and 2004; total local spending slowed from trend in 2003 but then increased again in III. State Spending, Local Spending, and the 2001 Recession The previous section focused on spending patterns at the aggregate, or national, level. We turn now to how those patterns vary within and across states, focusing first on the changes in real perstudent spending by state plus local governments, state governments only, and local governments only. We then turn to an analysis of the deviations from trend. Again, we exclude federal revenue. Table 3 (p. 642) shows the relationship between state and local changes from 2001 in real revenues per student. Appendix tables A-1 (p. 649), A-2 (p. 650), and A-3 (p. 651) present the changes for state plus local, for state, and for local, respectively, by state. We categorized the states into four groups: Group 1 (G1) are those states for which both state and local spending per student fell; Group 2 (G2) are those states for which the state spending fell but local spending increased; Group 3 (G3) are those states for which state spending increased but local per student, right scale School year ended spring $8,000 7,500 7,000 6,500 6,000 5,500 5,000 4,500 4,000 Per Student spending fell; and Group 4 (G4) are those states for which both state and local spending increased. In half of the states, both state and local spending per student increased between 2001 and However, there were only 17 states for which the change between 2001 and 2004 was positive for both state and local spending. What is striking is that the change in local spending from 2001 to 2004 was negative in only two states; that is, local spending per student in 2004 increased substantially in many states to offset the reductions in 2002 and How did the change in spending per student differ across regions? Table 4 (p. 643) shows the change in average real spending per student between 2001 and 2004 by region, where the average is the unweighted average across states within the region. In all regions, average real spending per student increased at the local level, while it decreased at the state level. The increase in local spending in the Northeast was substantially larger than for other regions. However, there was little difference across regions in the decrease in state spending per student. The net result was that average real state plus local spending per student increased in all regions, but only by a very small amount in the South and West. 640 State Tax Notes, May 28, 2007

5 Total, Billions $ Figure 3. State Education Revenue (50 states and D.C., Real 2000 dollars, with trend) Legend: Expenditures per student Total expenditures on K-12 education (in billions) Trend lines Recession periods 1994 That pattern can be seen in maps 1, 2, and 3, which show the percentage change in real spending per student between 2001 and 2004 for state plus local, for state, and for local governments, respectively. There were 12 states for which real state and local spending per student fell between 2001 and 2004 (Map 1, p. 643). None of those states was in the Northeast, which is consistent with the relatively large increase in average spending per student seen in Table 4. As seen in Map 2 (p. 644), no state in the Midwest and only one in the Northeast had a decline in state government spending per student, which is also consistent with the relatively smaller decreases in average spending per student reported in Table 4. There were only two states in which local spending per student fell, one in the Midwest and one in the West (Map 3, p. 645). The analysis of actual changes in revenue per student does not consider the historic pattern of spending on K-12 education in a state. Because the trend in per-student spending varies substantially across states, a decrease in spending over the previous year for a state in which revenue per student had been increasing rapidly suggests a more sizable fiscal effect than for a state for which revenue per per student, right scale School year ended spring $4,500 4,000 3,500 3,000 2,500 2,000 student had not been increasing. Thus, we again consider deviations from trend. Appendix tables A4 (p. 655), A5 (p. 656), and A6 (p. 657) present the deviations for the three divisions of government and the three years. Those tables also contain the coefficient from the log-linear trend regression along with the level of significance and R 2. In all regions, average real spending per student increased at the local level, while it decreased at the state level. Per Student As before, we categorized the states into the same four groups: Group 1 (G1) is those states for which both state and local spending per student fell below trend; G2 is those states for which the state spending fell below trend but local spending increase from trend; G3 is those states for which state spending increased from trend but local spending fell below trend; and G4 is those states for which both state and local spending increased from trend. Table 5 (p. 644) shows those distributions for 2002, 2003, and There is some slight tendency for local spending per State Tax Notes, May 28,

6 Total, Billions $ Figure 4. Local Education Revenue (50 states and D.C., Real 2000 dollars, with trend) Legend: Expenditures per student Total expenditures on K-12 education (in billions) Trend lines Recession periods 1994 per student, right scale School year ended spring $4,500 4,000 3,500 3,000 2,500 2,000 Table 2. Deviations From Trend (in percentages) Total Revenue Total Revenue per Student Year State plus Local State Local State plus Local State Local % 0.2% 0.4% 1.1% 0.8% 1.4% % -3.3% -1.0% -1.5% -2.8% 0.1% % -8.8% 6.0% -1.3% -8.3% 7.5% Source: Calculations by authors. Per Student Table 3. Distribution of Percentage Change From 2001 in Real Expenditures per Student State Negative and Local Negative (G1) State Negative and Local Positive (G2) State Positive and Local Negative (G3) Both State and Local Positive (G4) Total State Negative Total Local Negative Source: Calculations by authors. 642 State Tax Notes, May 28, 2007

7 Table 4. Change in Real per-student Spending, , by Region (in dollars) Region Local State State and Local South $ $ $98.08 West Midwest Northeast Source: Calculations by authors. Map 1. Percentage Change in State and Local Real Spending per Student, State and Local Percent Change, > = 10% (7) 3% to 5% (12) Less than 0% (12) 5% to 10% (8) 0% to 3% (11) student to have a positive (negative) deviation when the state has a negative (positive) deviation; 27 states fall into that category for 2002, 26 for 2003, and 33 for Note that, because predicted spending based on the trend is increasing over time in nearly all cases, deviations from trend will increase over time unless actual spending increases more rapidly than the trend. Table 6 (p. 654) shows the distribution of the number of years for which state or local spending was below trend. Deviations from trend were more likely to be negative for state spending than for local spending. If there were negative deviations, it was more likely that they were observed for more than one year, with three years being the mode. Nearly half of the states had a negative deviation from trend for all three years. Deviations in state spending were more likely to be negative in 2004 (34 states), while deviations for local spending were more likely to be negative in 2003 (27 states). It is those deviations from trend that we focus on in much of the subsequent discussion. State Tax Notes, May 28,

8 Table 5. Distribution of Deviations From Trend in Real Expenditures per Student State Negative and Local Negative (G1) State Negative and Local Positive (G2) State Positive and Local Negative (G3) Both State and Local Positive (G4) Total State Negative Total Local Negative Source: Calculations by authors. Map 2. Percentage Change in State Real Spending per Student, State Percent Change, > = 5% (10) -3% to 0% (7) Less than -7.5% (11) 0% to 5% (9) -7.5% to -3% (13) It is of some interest to explore more formally how local spending changed as a result of changes in state spending, using these deviations from trend. In particular we address the question whether local spending offset decreases in state spending from trend; that is, if the state deviation from trend is larger and more negative, will the local deviation from trend be larger and more positive to counter the state changes? Why might such a pattern emerge? The median voter theory suggests that voters will select the desired total spending per student, and the split between state and local will be determined separately (Thomas 2000). Thus, if state spending on education is reduced or falls below trend, school district voters might decide to offset at least part of that decrease with an increase in local spending. 644 State Tax Notes, May 28, 2007

9 Table 6. Years Below Trend, State None 1 Year 2 Years 3 Years Row Total None Year Local 2 Years Years Column Total Source: Calculations by authors. Map 3. Percentage Change in Local Real Spending per Student, Local Percent Change, > = 20% (9) 6% to 10% (11) Less than 0% (2) 10% to 20% (14) 0% to 6% (14) To examine that pattern for those states with an overall state and local negative deviation in 2004, we calculate the average dollar deviation for both state spending per student and for local spending per student. (Recall that the average local spending deviation in 2004 was positive.) We then calculate the local deviation as a percent of the state deviation. The increase in local spending per student above trend was 42.3 percent of the decrease from trend in state spending per student, where these figures represent the unweighted average of the decrease in state spending from trend that was replaced by increased local spending per student above trend. That is a significant replacement rate. To further investigate that pattern, we regressed the local deviation (in dollars) against the state deviation for each year (plus a constant), using both percent and dollar deviations. Because Alaska and Hawaii are special cases, we excluded them from the analysis. The coefficient on the state deviation variable is negative in all cases, suggesting that the more negative the state deviation, the more likely State Tax Notes, May 28,

10 Map 4. Percentage Deviation in State Plus Local Spending per Student, 2004 the local deviation is to be positive. However, the coefficients are not all statistically significant, though they are (at the 5 percent level) for 2004 for both total and per-student spending. Consider the coefficient on per-student deviations for 2004, that is, That estimate suggests that, for each dollar that state spending was below trend in 2004, local spending per student on average increased by 44 cents. That replacement of 44 percent is close to the unweighted average replacement rate of 42.3 percent reported above. 7 7 Note that local school districts face different tax prices for increasing spending on K-12 because of the nature of the state education finance system. If locally raised money at the margin is matched with state funds, the local tax price of increasing spending per student is less than $1; conversely, if the state takes a percentage of locally raised school revenue above a certain level, the local tax price is greater than $1. As a result, the local district may be more or less inclined (or able) to increase local revenue, depending on the state education finance system. Hoxby (2001) calculated the tax price for each state for Unfortunately, many states changed State and Local Deviation From Trend, 2004 > = 5% (10) -3.5% to 0% (10) Less than -8% (9) 0% to 5% (12) -8% to -3.5% (9) We also consider how the deviations from trend for 2004 vary across regions. Table 7 is equivalent to Table 4 except that Table 7 presents deviations from trend rather than changes in actual spending. The patterns in the two tables are somewhat similar. The deviation in local spending is positive and the deviation for state spending is negative as in Table 4, but state plus local spending per student was below trend in three of the regions. The major difference in the magnitude of the deviations as compared with the change reported in Table 4 is for the Midwest. Note that the state plus local deviation is positive for the West. Maps 4, 5, and 6 show the variation across their school finance system during the 1990s. However, in regressions of local dollar deviations from trend-predicted expenditures, we included her minimum tax price along with the state dollar deviation for Including the tax price (inverted) did not change the coefficient on state deviations; the coefficient on the (inverted) tax price was statistically significant, but its sign ran counter to the expected positive sign. (Footnote continued in next column.) 646 State Tax Notes, May 28, 2007

11 Map 5. Percentage Deviation in State Spending per Student, 2004 states for the 2004 state and local, state, and local per-student deviations. The patterns are similar to those seen in maps 1 through 3. IV. Conclusions Our analysis suggests that the recession of 2001 had a significant though somewhat variable impact on state and local government education spending. We found that real total state and local government K-12 spending was slightly above trend in 2002 and that total state and local spending fell significantly below trend in 2003 and in However, although education spending at the state level remained well below trend in 2004, spending at the local level State Deviations From Trend, 2004 >=5%(9) -7% to 0% (11) Less than -15% (14) 0% to 5% (6) -15% to -7% (10) recovered in 2004 and was significantly above trend in that year. Also, there is some slight tendency for local spending per student to have a positive (negative) deviation from trend when the state has a negative (positive) deviation from trend. References Alm, James, Robert Buschman, and David L. Sjoquist. The 2001 Recession, State and Local Government Fiscal Problems, and Government Education Spending. Working paper, Fiscal Research Center, Andrew Young School of Policy Studies, Georgia State University (2007). Table 7. Deviation in 2004, by Region (in dollars) Region Local State State and Local South $ $ $ West Midwest Northeast Source: Calculations by authors. State Tax Notes, May 28,

12 Map 6. Percentage Deviation in Local Spending per Student, 2004 Hoxby, Caroline M. All School Finance Equalizations Are Not Created Equal: Marginal Tax Rates Matter. The Quarterly Journal of Economics 116 (November 2001): Holahan, John, Teresa A. Coughlin, Randall R. Bovbjerg, Ian Hill, Barbara A. Ormond, and Stephen Zuckerman. State Responses to 2004 Budget Crises: A Look at Ten States. Washington, D.C.: The Urban Institute (February 2004). Johnson, Nicholas, Jennifer Schiess, and Joseph Llobrera. State Revenues Have Fallen Dramatically: Tax Increases So Far Failed to Fill the Gap. Washington: Center on Budget and Policy Priorities (Oct. 22, 2003). Kalambodkidis, Laura, and Andrew Reschovsky. States Responses to the Budget Shortfalls of Challenge 48 (1) (January-February 2005): Ladd, Helen F. How School Districts Respond to Fiscal Constraint in Selected Papers in School Local Deviation From Trend, 2004 > = 10% (13) 0% to 5% (11) Less than -5% (8) 5% to 10% (9) -5% to 0% (6) Finance, 1996, William J. Fowler Jr. (ed). Washington: National Center for Education Statistics (1997): McNichol, Elizabeth, and Makeda Harris. Many States Cut Budgets as Fiscal Squeeze Continues. Washington: Center on Budget and Policy Priorities (Apr. 26, 2004). National Education Association. Rankings & Estimates: Rankings of States 2004 and Estimates of School Statistics 2005 (2006), available at Reschovsky, Andrew. The Impact of State Government Fiscal Crises on Local Governments and Schools. State and Local Government Review 36 (2) (Spring 2004): Thomas, Mary Kathleen. The State Share of Revenue for Public Education: Measuring Its Magnitude and Effect on Total Per Pupil Spending in Primary and Secondary School Districts. Ph.D. dissertation, Atlanta: Georgia State University (2000). 648 State Tax Notes, May 28, 2007

13 Appendix Table A-1. Percent Change From 2001 in Real State and Local Spending per Student Actual Real per-student SL Spending Percent Change From 2001 State Alabama $5, , $5, % -2.0% 4.4% Alaska 8, , , % -1.9% 1.3% Arizona 5, , , % 9.5% 4.3% Arkansas 5, , , % 2.8% 6.7% California 7, , , % -0.6% -4.7% Colorado 6, , , % 6.1% 9.3% Connecticut 10, , , % 0.5% 1.1% Delaware 8, , , % -1.2% 2.3% Florida 5, , , % -6.5% 2.0% Georgia 7, , , % -2.0% -4.2% Hawaii 8, , , % 16.8% 13.1% Idaho 5, , , % -3.4% -4.6% Illinois 7, , , % -3.9% 1.3% Indiana 7, , , % -21.6% -0.5% Iowa 6, , , % 2.0% 2.8% Kansas 6, , , % 3.0% 13.6% Kentucky 5, , , % -0.9% 1.6% Louisiana 5, , , % 2.6% 4.0% Maine 8, , , % 5.0% 5.3% Maryland 8, , , % 1.6% 4.6% Massachusetts 9, , , % 7.2% 3.2% Michigan 8, , , % -2.0% -1.9% Minnesota 8, , , % -1.1% 0.9% Mississippi 4, , , % 4.6% 10.8% Missouri 6, , , % 0.5% 3.6% Montana 6, , , % -1.2% 4.7% Nebraska 7, , , % 2.5% 8.3% Nevada 6, , , % -1.4% 3.8% New Hampshire 7, , , % 6.8% 12.9% New Jersey 11, , , % 6.1% 11.3% New Mexico 6, , , % 2.6% 5.0% New York 10, , , % 2.1% 6.0% North Carolina 5, , , % -10.5% -9.5% North Dakota 5, , , % 3.5% 14.4% Ohio 8, , , % 1.8% 4.4% Oklahoma 5, , , % -6.3% -0.6% Oregon 7, , , % -8.5% 1.1% Pennsylvania 8, , , % 1.6% 5.0% Rhode Island 9, , , % 3.9% 6.7% South Carolina 6, , , % -5.8% -3.3% South Dakota 5, , , % -3.4% 4.9% Tennessee 5, , , % -6.7% 0.4% Texas 6, , , % 0.3% -1.8% Utah 5, , , % -3.7% -3.8% Vermont 9, , , % 5.0% 7.5% Virginia 7, , , % -0.2% 2.1% Washington 6, , , % -1.2% -0.4% West Virginia 7, , , % 1.8% 3.5% Wisconsin 8, , , % -1.7% -1.1% Wyoming 8, , , % 14.6% 12.5% State Tax Notes, May 28,

14 Table A-2. Percent Change From 2001 in Real State Spending per Student Actual Real per-student State Spending Percent Change From 2001 State Alabama $3, $3, $3, % -3.4% -5.7% Alaska 5, , , % -0.8% 0.6% Arizona 3, , , % 23.6% 12.7% Arkansas 3, , , % 0.7% -3.0% California 4, , , % -3.1% -12.2% Colorado 3, , , % 10.4% 9.7% Connecticut 4, , , % -3.9% -7.9% Delaware 6, , , % -4.2% -3.8% Florida 3, , , % -14.4% -10.5% Georgia 4, , , % -1.5% -9.1% Hawaii 8, , , % 16.6% 10.8% Idaho 3, , , % -5.1% -9.0% Illinois 2, , , % -4.8% -1.0% Indiana 4, , , % -10.9% -6.5% Iowa 3, , , % -2.2% -7.0% Kansas 4, , , % -0.8% -5.0% Kentucky 4, , , % -2.2% -2.9% Louisiana 3, , , % 4.0% 3.7% Maine 4, , , % 2.1% -1.7% Maryland 3, , , % 4.5% 3.5% Massachusetts 4, , , % 1.6% -4.1% Michigan 6, , , % -3.2% -7.5% Minnesota 5, , , % 22.4% 13.9% Mississippi 3, , , % 3.4% 7.6% Missouri 2, , , % -2.7% -7.4% Montana 3, , , % -0.4% -2.1% Nebraska 2, , , % 2.5% -2.3% Nevada 2, , , % 6.3% 5.8% New Hampshire 4, , , % 1.9% -1.1% New Jersey 5, , , % 11.1% 13.9% New Mexico 5, , , % 5.3% 4.3% New York 5, , , % 2.1% 0.4% North Carolina 4, , , % -11.7% -13.6% North Dakota 2, , , % -0.2% 7.4% Ohio 4, , , % 5.9% 5.6% Oklahoma 3, , , % -10.7% -11.0% Oregon 4, , , % -15.6% -7.7% Pennsylvania 3, , , % -0.2% -0.7% Rhode Island 4, , , % 4.1% 4.3% South Carolina 4, , , % -14.2% -18.4% South Dakota 2, , , % -3.7% 2.7% Tennessee 2, , , % -2.7% -4.3% Texas 3, , , % -1.7% -10.3% Utah 3, , , % -5.7% -8.5% Vermont 7, , , % 1.8% 1.2% Virginia 3, , , % -5.7% -7.0% Washington 4, , , % -1.6% -4.9% West Virginia 5, , , % 2.8% 2.9% Wisconsin 4, , , % 0.0% -3.6% Wyoming 4, , , % 16.2% 16.3% 650 State Tax Notes, May 28, 2007

15 Table A-3. Percent Change From 2001 in Real Local Spending per Student Actual Real per-student Local Spending Percent Change From 2001 State Alabama $1, $1, $2, % 1.4% 28.2% Alaska 2, , , % -4.6% 3.1% Arizona 2, , , % -4.6% -4.2% Arkansas 1, , , % 7.2% 26.4% California 2, , , % 4.8% 11.2% Colorado 3, , , % 2.4% 9.0% Connecticut 5, , , % 3.7% 7.5% Delaware 2, , , % 6.6% 18.6% Florida 2, , , % 3.4% 17.7% Georgia 3, , , % -2.5% 1.5% Hawaii % 62.0% 405.0% Idaho 1, , , % 0.3% 4.8% Illinois 4, , , % -3.3% 2.6% Indiana 3, , , % -36.6% 7.8% Iowa 3, , , % 7.2% 15.2% Kansas 2, , , % 10.7% 51.6% Kentucky 1, , , % 1.9% 11.3% Louisiana 2, , , % 0.9% 4.4% Maine 4, , , % 7.9% 12.2% Maryland 4, , , % -0.4% 5.3% Massachusetts 5, , , % 12.1% 9.7% Michigan 2, , , % 1.1% 11.9% Minnesota 2, , , % -46.0% -24.0% Mississippi 1, , , % 7.0% 17.2% Missouri 4, , , % 2.7% 11.6% Montana 2, , , % -2.2% 13.6% Nebraska 4, , , % 2.5% 15.4% Nevada 4, , , % -4.9% 2.9% New Hampshire 3, , , % 12.9% 30.3% New Jersey 6, , , % 2.1% 9.2% New Mexico , % -11.9% 8.5% New York 5, , , % 2.1% 11.4% North Carolina 1, , , % -7.1% 1.6% North Dakota 3, , , % 6.9% 20.8% Ohio 4, , , % -2.0% 3.2% Oklahoma 1, , , % 3.8% 23.3% Oregon 2, , , % 3.5% 16.0% Pennsylvania 4, , , % 2.8% 9.0% Rhode Island 4, , , % 3.8% 8.7% South Carolina 2, , , % 7.4% 20.4% South Dakota 3, , , % -3.2% 6.5% Tennessee 2, , , % -10.8% 5.2% Texas 3, , , % 2.0% 5.9% Utah 1, , , % 0.0% 5.0% Vermont 2, , , % 15.5% 27.9% Virginia 4, , , % 4.5% 9.8% Washington 2, , , % 0.0% 10.5% West Virginia 2, , , % -0.3% 4.8% Wisconsin 3, , , % -4.0% 2.3% Wyoming 4, , , % 12.5% 7.6% State Tax Notes, May 28,

16 Table A-4. Total State and Local Revenue per Student (2000 constant dollars) Percent Over/Under Log-Linear Time Trend Predicted Trend Constant State β R Alabama *** $5, $5, $5, % -14.8% -13.1% Alaska *** , , , % 3.5% 9. 2% Arizona *** , , , % 8.1% 2.3% Arkansas *** , , , % -3.4% -2.3% California *** , , , % 4.0% -2.3% Colorado *** , , , % 4.3% 6.2% Connecticut *** , , , % 1.5% 0.9% Delaware *** , , , % -9.4% -8.9% Florida , , , % -7.5% 0.6% Georgia *** , , , % -6.7% -12.3% Hawaii *** , , , % 29.4% 23.8% Idaho *** , , , % -11.9% -16.2% Illinois *** , , , % -6.6% -4.2% Indiana *** , , , % -26.5% -9.4% Iowa *** , , , % -2.9% -4.5% Kansas *** , , , % -2.1% 6.3% Kentucky *** , , , % -8.3% -8.0% Louisiana *** , , , % -0.5% -1.9% Maine *** , , , % 3.1% 1.2% Maryland *** , , , % 2.9% 4.9% Massachusetts *** , , , % 7.5% 1.2% Michigan *** , , , % -6.2% -7.8% Minnesota *** , , , % -3.1% -3.4% Mississippi *** , , , % -5.6% -4.2% Missouri *** , , , % -4.1% -3.6% Montana *** , , , % -2.8% 1.7% Nebraska *** , , , % -2.7% 0.9% Nevada *** , , , % -5.9% -1.9% New Hampshire * , , , % 8.5% 14.2% New Jersey , , , % 7.7% 12.7% New Mexico *** , , , % 1.0% 0.5% New York *** , , , % 3.6% 6.2% North Carolina *** , , , % -12.8% -13.9% North Dakota *** , , , % 1.9% 10.4% Ohio *** , , , % -1.2% -1.8% Oklahoma * , , , % -4.2% 0.7% Oregon *** , , , % -9.6% -1.7% Pennsylvania *** , , , % 0.3% 2.8% Rhode Island *** , , , % 1.3% 2.0% South Carolina *** ,887,.04 6, , % -7.3% -7.9% South Dakota *** , , , % -8.6% -3.4% Tennessee *** , , , % -12.9% -9.7% Texas *** , , , % -3.0% -7.1% Utah *** , , , % -11.7% -14.8% Vermont *** , , , % 8.0% 9.5% Virginia *** , , , % -2.2% -1.7% Washington *** , , , % -1.7% -1.5% West Virginia *** , , , % -4.8% -5.3% Wisconsin *** , , , % -4.9% -6.1% Wyoming *** , , , % 15.1% 11.7% 652 State Tax Notes, May 28, 2007

17 Table A-5. State Revenue per Student (2000 constant dollars) Percent Over/Under Log-Linear Time Trend Predicted Trend Constant State β R Alabama *** $3, $3, $3, % -17.4% -22.5% Alaska *** , , , % 6.0% 11.2% Arizona *** , , , % 19.3% 7.4% Arkansas *** , , , % -6.7% -12.4% California , , , % 7.0% -4.5% Colorado *** , , , % 4.8% 2.5% Connecticut ** , , , % -0.9% -5.7% Delaware *** , , , % -9.7% -11.9% Florida , , , % -15.7% -12.3% Georgia *** , , , % -6.5% -16.6% Hawaii *** , , , % 29.2% 21.2% Idaho *** , , , % -13.6% -20.1% Illinois *** , , , % 2.4% 3.0% Indiana *** , , , % -14.1% -12.4% Iowa *** , , , % -11.2% -18.3% Kansas *** , , , % -17.4% -24.9% Kentucky *** , , , % -7.3% -9.0% Louisiana *** , , , % 2.9% 1.1% Maine *** , , , % 3.5% -1.4% Maryland *** , , , % 2.0% -0.3% Massachusetts *** , , , % -5.5% -15.4% Michigan (1995) * , , , % -8.8% -14.4% Minnesota *** , , , % 19.3% 7.1% Mississippi *** , , , % -6.6% -6.6% Missouri *** , , , % -9.3% -15.8% Montana (1993) , , , % 2.1% -0.5% Nebraska *** , , , % -4.2% -11.2% Nevada ** , , , % 8.1% 9. 5% New Hampshire (2000) , , , % -2.0% -6.0% New Jersey , , , % 14.9% 17.0% New Mexico *** , , , % 3.5% -0.1% New York *** , , , % 9.0% 5.0% North Carolina *** , , , % -16.0% -20.0% North Dakota *** , , , % -0.1% 6.5% Ohio *** , , , % 5.2% 1.5% Oklahoma *** , , , % -11.2% -12.7% Oregon (1996) *** , , , % -36.4% -35.7% Pennsylvania , , , % 1.7% 1.4% Rhode Island *** , , , % 1.7% -0.6% South Carolina *** , , , % -17.2% -24.6% South Dakota *** , , , % -18.0% -18.1% Tennessee *** , , , % -15.1% -19.4% Texas *** , , , % -6.3% -16.8% Utah *** , , , % -17.4% -23.1% Vermont (1999) * , , , % 7.2% 7.7% Virginia *** , , , % -2.7% -7.9% Washington *** , , , % 1.0% -1.6% West Virginia *** , , , % -1.8% -3.1% Wisconsin (1997) *** , , , % -7.5% -13.5% Wyoming *** , , , % 14.9% 13.0% State Tax Notes, May 28,

18 Table A-6. Local Revenue per Student (2000 constant dollars) Percent Over/Under Log-Linear Time Trend Predicted Trend Constant Satate β R Alabama *** $1, $1, $2, % -9.0% 9.3% Alaska *** , , , % -4.6% 1.5% Arizona , , , % -2.8% -2.3% Arkansas *** , , , % 5.0% 20.9% California *** , , , % -2.8% -0.2% Colorado ** , , , % 3.7% 9.5% Connecticut *** , , , % 2.5% 4.8% Delaware *** , , , % -7.3% -0.5% Florida , , , % 3.2% 17.4% Georgia *** , , , % -6.3% -6.6% Hawaii % 87.9% 481.0% Idaho *** , , , % -8.5% -8.0% Illinois *** , , , % -10.7% -7.4% Indiana *** , , , % -42.1% -4.6% Iowa *** , , , % 9.1% 15.5% Kansas *** , , , % 26.3% 79.1% Kentucky *** , , , % -12.0% -8.7% Louisiana *** , , , % -5.9% -6.8% Maine *** , , , % 2.2% 3.0% Maryland ** , , , % 3.3% 8.5% Massachusetts , , , % 17.0% 14.2% Michigan (1995) *** , , , % -4.3% 3.4% Minnesota , , , % -46.7% -25.2% Mississippi *** , , , % -2.9% 1.6% Missouri *** , , , % 0.4% 6.2% Montana (1993) , , , % -8.2% 4.5% Nebraska *** , , , % -1.0% 9.9% Nevada *** , , , % -13.0% -8.3% New Hampshire (2000) , , , % 21.1% 39.5% New Jersey , , , % 2.4% 9.6% New Mexico *** , % -10.8% 5.9% New York ** , , , % -0.5% 8.2% North Carolina *** , , , % -3.4% 4.3% North Dakota *** , , , % 3.2% 13.0% Ohio *** , , , % -6.9% -4.8% Oklahoma , , , % 13.4% 35. 2% Oregon (1996) * , , , % 23.3% 42.4% Pennsylvania *** , , , % -1.2% 3.0% Rhode Island *** , , , % 1.1% 4.4% South Carolina *** , , , % 8.3% 19.2% South Dakota ** , , , % -3.5% 5.5% Tennessee *** , , , % -10.3% 1.7% Texas *** , , , % 0.4% 2.3% Utah *** , , , % -0.2% 2.2% Vermont (1999) ** , , , % 28.6% 39.2% Virginia , , , % -1.9% 2.9% Washington *** , , , % -9.5% -4.1% West Virginia *** , , , % -12.7% -12.0% Wisconsin (1997) *** , , , % -4.2% 0.9% Wyoming , , , % 15.8% 10.1% 654 State Tax Notes, May 28, 2007