Monitoring local government fiscal health: Michigan's new 10 point scale of fiscal distress.
Michigan has two laws (Public Act 70 of 1990 and Public Act 34 of 2001) intended to provide an early warning of fiscal distress. Together, these statutes contain 30 triggers that, if tripped, initiate a review process that can ultimately result in a state takeover of local government finances. In some cases, however, state officials were not alerted until serious difficulties had already occurred. To remedy this situation, the Michigan Department of Treasury contracted with the Institute for Public Policy and Social Research at Michigan State University to identify new measures that could better predict fiscal distress. (2) This article outlines the composite devised for Michigan and how it builds on previous work, including a composite measure proposed by Ken Brown and published in the December 1993 issue of Government Finance Review. (3)
SHORTCOMINGS OF MICHIGAN'S INDICATORS
Michigan's set of 30 indicators was viewed as ineffective on many fronts:
* Frequent, publicly available, and uniformly collected data do not exist for many of the triggers.
* Most of the triggers do not have theoretical validity, but instead focus on technical violations or requests for review.
* There is no degree of proportion reflected by the 30 triggers in the two acts. If a government is violating lust one of these conditions (by perhaps being a month late in delivering a financial report), then it appears to be as technically "fiscally distressed" as one that is in violation of several and more serious triggers. This provides little ability for early warning, since there is no sense of gradation in the level of distress that a government is experiencing.
* In terms of early warning, Michigan's indicators do not appear reliable from either a Type I (false positive) or Type II (false negative) standpoint. False positives could easily occur, and units headed for trouble could routinely escape detection even as they are headed for a fiscal emergency.
* Several, if not most, of Michigan's indicators are more suited to defining rather than predicting fiscal distress. By the time many of these triggers are violated, the government is in serious fiscal straits.
Although several other measures of fiscal distress were available, many suffered from limitations that prohibited their use for Michigan's purposes. (4) One of the measures that most closely approximated Michigan's needs was that presented by Brown 10 years ago in Government Finance Review. (5) In the few states that monitor and evaluate local government financial condition through ratio indicators, Brown's index has been explicitly acknowledged as influencing their system.
BROWN'S 10-POINT TEST
Brown suggested that 10 ratio measures be computed, equally weighted, and aggregated to provide an overall picture of a government's financial condition. Exhibit 1 shows the formulas for computing each of the 10 ratios. The ratios are computed for all of the local governments in a state and then assigned to quartiles. Governments receive points for each ratio depending on the quartile in which the ratio falls: two points for each ratio in Quartile 4 (75 to 100 percentile), one for Quartile 3(50 to 75 percentile), zero for Quartile 2(25 to 50 percentile), and minus one for Quartile 1 (0 to 25 percentile). Under this grading system, governments can earn as many as 20 points or as few as minus 10.
Brown's financial condition test provided a positive innovation in that it attempted to provide an overall assessment of local government financial condition using a straightforward test based on generally available data. While this test most closely approximated Michigan's needs for an early warning system, numerous practical and theoretical difficulties required that the state seek a different tool. Here we discuss some of the limitations of Brown's model.
Relativity. Brown's test rewards or punishes governments on a relative rather than on an absolute basis for the individual indicators. Governments in the top quartile are always rewarded and those in the bottom quartile are always penalized, regardless of their absolute merit. For example, even if all governments had large unreserved general fund surpluses (generally a desirable outcome), the Brown system would still penalize those with the smallest surpluses. Similarly, if all governments were running operating deficits, those with the smallest deficits would be awarded two points. A system that rewards or penalizes governments based on how their actions relate to fiscal distress in an absolute sense is more appropriate than Brown's 10-point test.
Lack of Multi-Year Comparability. If the relative nature of individual indicators is a problem, then logically the same difficulty holds for the composite score that results from summing the scores of those individual indicators. Such a system could lead to a more permanent set of fiscally distressed governments. Even if all governments were doing well on all 10 ratios (in an absolute sense), some would probably score poorly in the aggregate, since someone has to be in the bottom quartiles.
The relative nature of the aggregate score becomes even more problematic when it is used to determine whether a state takeover of local finances is warranted. Because scores are a function of relative performance for a given year, they are not directly comparable across years. This makes it difficult to discern if a local government's fiscal position has actually improved over time. It also complicates the process of identifying a score that would consistently justify some level of state intervention.
There are no explicit year-to-year implications reflected in Brown's measures. Instead, each government is evaluated solely on measures reported for a single year. The test would be better served by an expanded analytical timeframe that allowed for a more complete evaluation of financial condition. For example, while an operating deficit for one fiscal year is certainly an indicator of fiscal stress, consecutive deficits or increasing deficits would indicate a much more serious problem. A longer analytical timeframe would also eliminate the problem of placing too much weight on idiosyncrasies of a particular year. Brown's single year analysis does not allow for this.
Lack of Social, Economic Indicators. Although Brown's index appropriately incorporates several balance sheet-based measures, it does not include any social or economic measures. Since theory suggests that fiscal distress may be the result of tax base shifts, measures reflecting these types of trends should be included in the model.
Unsuitable Indicators. At least three of the 10 variables are not suitable indicators of fiscal distress. These include per capita revenues, per capita direct long-term debt, and general fund revenues from own sources as a percentage of total general fund revenues. Per capita revenue is an inappropriate measure because both wealthy and poor communities may have relatively high per capita revenues. In Michigan, the governments that have historically experienced severe fiscal distress have placed among the highest in per capita revenues.
Per capita direct long-term debt has the same deficiency as per capita revenue, since it does not distinguish among governments. Governments that have experienced fiscal distress are indeed likely to have high per capita debt, but wealthy jurisdictions also score quite high on a per capita basis.
Total general fund revenues from own sources as a percentage of total general fund revenues does not appear to be related to fiscal distress in Michigan. Although Michigan's townships are the most reliant on other sources for revenues, very few townships experienced serious fiscal difficulties in the time period we examined. The cities that did experience severe fiscal trouble did not appear to score atypically from other municipalities on this variable, and their reliance on other governments was routinely less than the level of townships.
All or Nothing. Another weakness of the Brown approach is that ratios for all units must be computed before any determination can be made about the relative fiscal health of a single unit. Should a state desire to conduct targeted oversight for only some jurisdictions, it would have to compute the scores for all similar governments in the state and then assign points based on relative performance. Requiring all governments to be measured before a single government can be evaluated may not be a wise use of resources when alternatives that rely on objective rather than relative performance could be used. This reliance on relative performance is further complicated by the possibility that several units may not submit their reports or audits in a timely manner. How to account for this missing data is not clear, and it is especially problematic given that late reporting appears to be more common among distressed governments than their fiscally sound counterparts.
A NEW MEASURE OF FISCAL DISTRESS
In addition to the concerns identified in the previous section was the very practical problem that Michigan simply did not have the kind of data needed to construct some of the indicators in the manner Brown described. Given the weaknesses of both Michigan's own indicators and Brown's 10-point test, it became clear that a new measure of fiscal distress would have to be constructed.
We constructed a 10-point scale of fiscal distress that is a composite of nine variables. For each of the nine variables, we established a performance standard to be used in grading local government financial condition. (6) Some of these standards were based on the distribution of a sample of Michigan localities on these variables. So while the performance standards might vary from state to state, the approach can easily be adopted by different states. Jurisdictions that do not meet the standard for a given variable are penalized with points, while those that do meet the standard receive no points. The result is a system in which a score of "10" indicates severe fiscal distress and a score of "0" indicates little or no distress. Exhibit 2 identifies each of the nine indicators and the corresponding performance standard.
We applied this 10-point scale to 150 local governments in Michigan over an 11-year period from 1991 to 2001. The sample was randomly generated and supplemented with units that had been perceived as distressed by the state. The result of this application is reported in Exhibit 3, which lists all governments scoring 5 or more points.
The average score for all of the governments included in the sample was approximately 1.5. As expected, most of Michigan's localities are not experiencing fiscal difficulty. As Exhibit 3 shows, however, there are variations over the 10-year period. In 1994 and 1995, 13 governments scored a 5 or above; in 1996, that number had dropped to only five.
Importantly, the 10-point scale performed fairly well in identifying the governments that had been classified by the state as distressed. During the time period examined, Michigan had appointed review teams to assess the finances of Highland Park, Hamtramck, and Flint. Highland Park's review team was established in 1996, disbanded in 1999, and reinstated in 2000. Hamtramck's review team was established in 2000, and Flint had a review team appointed in 2001. In each of these cases, the 10-point scale pointed to fiscal distress well before the review teams were appointed. Several of the other localities registering relatively high scores on the 10-point scale have historically teetered on the edge of fiscal distress, including Benton Harbor, Ecorse, River Rouge, and Detroit.
To help state officials identify fiscally troubled local governments before the problem becomes a crisis, the 10-point scale can be used to classify local governments into one of four categories: fiscally healthy (0-4 points), fiscal watch (5 points), fiscal warning (6-7 points), and fiscal emergency (8-10 points). States can attach remedial actions to each category, ranging from private notification to placement on a public "fiscal watch list" to mandatory consideration of a review team. While these categories and consequences are merely suggestive, they demonstrate how the 10-point scale can be used to monitor the financial position of local governments.
While Brown's 10-point test of local government financial condition has been useful to some policymakers, there are practical and theoretical difficulties that limit its usefulness. The new composite model outlined here and tested using a sample of Michigan local governments builds on and improves that work, adding components that will allow states to recognize local fiscal difficulties before they become fiscal emergencies. The Michigan Department of Treasury is working internally and with local officials to collect information needed for this new measure of fiscal stress and to make both the index and its component parts available on the state Web site.
Exhibit I: Brown's 10 Key Ratios of Financial Condition
1. Total Revenues / Population
2. Total General Fund Revenues from Own Sources / Total General Fund Revenues
3. General Fund Sources from Other Funds / Total General Fund Sources
4. Operating Expenditures / Total Expenditures
5. Total Revenues / Total Expenditures
6. Unreserved General Fund Balance / Total General Fund Revenues
7. Total General Fund Cash and Investments / Total General Fund Liabilities
8. Total General Fund Liabilities / Total General Fund Revenues
9. Direct Long-Term Debt / Population
10. Debt Service / Total Revenues
Exhibit 2: 10-Point Scale of Fiscal Distress Indicator Performance Standard Population Growth If the government lost population, then (2 years) it is penalized one point. Real Taxable Value Growth If the government experienced negative (2 years) real growth, then it is penalized one point. Large Decrease in Real If growth in real taxable value is less Taxable Value (2 years) than -0.04, then the government is penalized one point. The level of -0.04 is approximately one standard deviation below the average two-year real growth rate for cities and villages and approximately 1.5 standard deviations below the township average. The standard used is closer to the city and village standard deviation because very few townships experienced fiscal distress. General Fund Expenditures If a city or village scores greater than as a Percentage of 0.05, or if a township scores greater Taxable Value than 0.01, then the government is penalized one point. This is the only variable for which we use a separate standard depending on the type of government. We did this because a half standard deviation in the "wrong direction" gives a standard of 0.05 for cities and villages and 0.01 for townships. General Fund Operating This indicator is calculated by Deficit subtracting general fund revenues from general fund expenditures for a given year and dividing the result by general fund revenues. If the result is less than -0.01, it is considered a nontrivial operating deficit and the government is penalized one point. Prior General Fund Governments are penalized one point for Operating Deficits each year in which they record an operating deficit. Thus, they can be penalized a total of three points for operating deficits--one for a current operating deficit and two for previous operating deficits. General Fund Balance as a If this ratio is less than 0.13, then Percentage of General the government is penalized one point, Fund Revenues Using a half standard deviation in the "wrong direction" as a benchmark (indicating a low fund balance). the resulting indicator threshold is about 0.13. Current or Previous Year Governments are penalized one point for Deficit in a Major Fund a current or previous year deficit in a major fund. For a definition of a major fund, see Stephen Gauthier, Governmental Accounting, Auditing, and Financial Reporting (Chicago: GFOA, 2001). General Long-Term Debt as If this ratio is greater than 0.06, then a Percentage of Real the government is penalized one point. Taxable Value The governments is our sample averaged 0.025 on this variable. Accordingly, one standard deviation in the "wrong direction" (high debt level) gives us a performance standard of about 6 percent. Exhibit 3: Historical Application of the New Composite Model 1993 Scores 9 Detroit 9 Pontiac 7 Flint 6 Benton Harbor 5 Ecorse 5 Saginaw 1994 Scores 7 Detroit 7 Pontiac 6 Flint 6 Highland Park 6 Ionia 6 Saginaw 5 Buena Vista Township 5 Ecorse 5 Manistique 5 Mount Clemens 5 Roosevelt Park 5 Royal Oak Township 5 Taylor 1995 Scores 7 Saginaw 6 Detroit 6 Gladstone 6 Hamtramck 6 Pontiac 5 Benton Harbor 5 Ecorse 5 Flint 5 Highland Park 5 Lansing 5 Manistique 5 Mount Clemens 5 Royal Oak Township 1996 Scores 7 River Rouge 5 Benton Harbor 5 Ecorse 5 Gladstone 5 Saginaw 1997 Scores 7 River Rouge 6 Benton Harbor 6 Buena Vista Township 6 Highland Park 5 Ecorse 5 Jackson 5 Royal Oak Township 1998 Scores 9 Highland Park 7 Buena Vista Township 7 Ecorse 6 Benton Harbor 5 Hampton Township 5 Hamtramck 5 Jackson 5 River Rouge 5 Royal Oak Township 1999 Scores 10 Highland Park 7 Hamtramck 6 River Rouge 5 Benton Harbor 5 Buena Vista Township 5 Ecorse 5 Flint 5 Jackson 5 Kalamazoo 5 Pontiac 2000 Scores 8 Flint 7 Benton Harbor 6 Ecorse 6 Kinross Township 5 Hamtramck 5 Highland Park 5 Newaygo 5 River Rouge 2001 Scores 9 Flint 7 Benton Harbor 7 Ecorse 6 Munising 6 Plainwell 5 Detroit 5 Kinross Township 5 Newaygo 5 Norway 5 Pontiac 5 Reading
(1.) Philip Kloha, Carol S. Weissert, and Robert Kleine, "Someone to Watch Over Me: State Practices in Monitoring Local Fiscal Conditions," prepared for the Midwest Political Science Association, April 3-5, 2003, Chicago.
(2.) Robert Kleine, Philip Kloha and Carol S. Weissert, "Fiscal Distress Indicators: An Assessment of Current Michigan Law and Development of a New 'Early-Warning' Scale for Michigan Localities" (East Lansing, MI: Institute for Public Policy and Social Research, 2002).
(3.) Ken W. Brown, "The 10-Point Test of Financial Condition: Toward an Easy-to-Use Assessment Tool for Smaller Cities," Government Finance Review 9 (December 1993): 6, 21-26; Ken W. Brown, "Trends in Key Ratios Using the GFOA Financial Indicators Database 1989-1993," Government Finance Review 12 (December 1996): 6,30-34.
(4.) Advisory Commission on Intergovernmental Relations, City Financial Emergencies.' The Intergovernmental Dimension (Washington D.C.: Government Printing Office, 1973); Municipal Finance Officers Association. Is Your City Heading for Financial Difficulty: A Guidebook for Small Cities and Other Governmental Units (Chicago: MFOA, 1978); Sanford M. Groves and Maureen G. Valente, Evaluating Financial Condition: A Handbook for Local Government (Washington D.C.: International City/County Management Association, 1994).
(5.) Brown (1993).
(6.) In some cases, the standard to use is intuitively straightforward, such as a declining tax base. In other cases, a clear benchmark is less obvious, and in these instances we adopt a standard deviation approach. This involves computing the average and standard deviation for the variable being considered. A benchmark is then set by taking the average value and combining it with a standard deviation in the "wrong direction." For example, the average for the debt variable is 0.025 and its standard deviation is 0.035. Since higher debt levels are undesirable, we added the standard deviation (3.5 percent) to the average (2.5 percent), giving a benchmark of 6 percent.
ROBERT KLEINE is a private consultant in Lansing, Michigan. He worked for the State of Michigan for 17 years, including 10 years as director of the Office of Revenue and Tax Analysis, and served as senior analyst for the Advisory Commission on Intergovernmental Relations. Mr. Kleine has also taught public finance as an adjunct professor at Michigan State University.
PHIL KLOHA is a graduate student in the Department of political science at Michigan State University He is interested in public policy, Congress, and state legislatures.
CAROL WEISSERT is professor of political science and director of the Institute for Public Policy and Social Research at Michigan State University Her research interests include federalism and intergovernmental relations, health policy, and legislative behavior Dr. Weissert holds a Ph.D. from the University of North Carolina
|Printer friendly Cite/link Email Feedback|
|Author:||Kleine, Robert; Kloha, Philip; Weissert, Carol S.|
|Publication:||Government Finance Review|
|Date:||Jun 1, 2003|
|Previous Article:||Risky business? Evaluating the use of pension obligation bonds.|
|Next Article:||The path to bond market efficiency: how increased retail distribution can lower borrowing costs.|