Does bankruptcy really matter? The solvency of municipal governments in the Chicago metropolitan region.
Rising pressures on government spending, revenue, and debt in the 2000s have worsened the financial condition of governments everywhere. The 2001 recession in the United States, the Great Recession worldwide, and methods of coping with these events have dramatically increased the deficits and fiscal obligations of governments at all levels. Some have predicted these conditions will create severe fiscal distress in governments in the US and abroad and will result in more governments filing for bankruptcy than in the past (Dubrow, 2009; Laughlin, 2005). Bankruptcy is considered to be the last resort for local governments that are financially insolvent and a valuable tool to help them return to solvency (Farmer, 2013). It protects them from creditors' demands and gives them the authority to adjust and restructure debt and contracts, which can lessen financial liabilities overall and given them more time to meet these obligations (Dubrow, 2009; Laughlin, 2005).
Despite the respite offered to local governments by bankruptcy, not all governments can file for this protection. Of the almost 90,000 local governments recognized in the US in 2007,1 only about 57% (in 24 states) were authorized either fully or conditionally by state government to file for bankruptcy under Chapter 9 of the federal Bankruptcy Code (15% had conditional authorization) (Spiotto, 2008). 2 Thus, findings from studies of bankruptcy in U.S. local governments are generalizable only to the population of local governments in states that authorize such filings. With respect to the population of insolvent governments in states that are authorized to file bankruptcy, the sample of bankrupt local governments in the US is further biased because many insolvent governments do not file for bankruptcy. The uncertainty and undesirability of the outcomes of bankruptcy give both governments and creditors incentives to avoid it and negotiate arrangements that help the government become solvent (Scorsone & Wright, 2013). Therefore, studying only bankrupt governments to understand why governments become insolvent is likely to confound state authorization and conditions surrounding the negotiation of arrangements to avoid bankruptcy with the causes of financial insolvency. It is more valid in this case to study insolvency directly in a general population of governments to determine its causes and possible remedies.
To study insolvency directly and determine its causes, one must first define and measure the concept in order to make accurate comparisons of solvent and insolvent governments and determine its prevalence. This study presents a model and strategy for determining the solvency or financial condition of local governments in the US and distinguishing between solvent and insolvent governments. We apply this conceptualization and methodology to all 265 suburban municipalities in the Illinois portion of Chicago metropolitan area from 1997 to 2010.3 Focusing this research on suburban governments greatly broadens the population of analysis from prior studies of fiscal crises in US governments that focus on central cities, which are much larger, less prevalent, and qualitatively different than suburban governments. This time period also allows us to compare these governments' financial condition during both fiscal bad times (the 2001 recession and the Great Recession) and fiscal good times (in the late 1990s).
We measure these governments' financial condition using a number of different indicators. First, we use a series of indicators that represent different dimensions and time frames of the phenomenon. These indicators are similar to what has been used by others in the field to differentiate between long-term and short-term solvency (Berne & Schramm, 1986; Nollenberger, Groves, & Valente, 2003; Standard & Poor's, 2012). We also define financial condition as the balance of factors within and across these dimensions (Clark & Ferguson, 1983; Ladd & Yinger, 1989), and we assume that these dimensions may be related in non-linear ways that preclude the ability to simply sum factors across dimensions to determine financial condition. Using this combined framework, we define relative insolvency on all indicators as whether a government is a member of the lowest two septiles of that indicator (for either good times or bad times) within the larger distribution. We also look at how the indicators are related to assess the frequency of insolvency among governments within the region and argue that the causes and risk of insolvency can be inferred from the joint distribution of governments on different dimensions of financial condition.
Although our research focuses on local governments in the state of Illinois, our model for conceptualizing financial condition and insolvency has international applicability as every country in the world likely has a population of insolvent local governments. For example, a 2004 World Bank paper (Intergovernmental Finance in Hungary: A Decade of Experience 1990-2000) that describes Hungary's experiences with the 1996 Act on Municipal Bankruptcy, also describes the insolvency of local governments in that country in a manner that is similar to what is conceptualized here. Clearly, municipal bankruptcy and insolvency are important issues for local governments worldwide, although specific factors of the dimensions that determine financial condition or solvency and the measurement of these factors are likely to vary significantly across countries and even states in the US. The next section explains how bankruptcy as defined in the US presents a limited view of solvency and why a broader focus is needed to accurately identify insolvent governments.
2. BANKRUPTCY, INSOLVENCY, AND FISCAL CRISES
2.1 Bankruptcy and Insolvency: A Rare Event with a Narrow Definition
The fact that so few local governments in the U.S. have filed for bankruptcy demonstrates how rare the event is and suggests that there are likely to be many insolvent governments that either have not or cannot file for bankruptcy. Between 1980 and 2012 only about 250 bankruptcy cases have been filed under Chapter 9, which was less than 0.3% of all local governments in the US in 2007, and most of these governments are utilities and special districts, (Deal, Heier, & Kamnikar, 2013). In fact, general-purpose governments represented only 17% of filings between 1980 and 2006 (Spiotto, 2008). Additionally, there are far more defaults on debt by local governments in the U.S. than bankruptcies,4 which demonstrates that insolvency is a necessary but not nearly sufficient condition for the occurrence of bankruptcy in the U.S. Governments that are insolvent or near insolvency according to the U.S. bankruptcy code and their creditors have strong incentives to negotiate a restructuring of government debt and other forms financial assistance to avoid filing for bankruptcy (Gillette, 2012).
Current bankruptcy law in the US requires local governments to satisfy five criteria in order to file for Chapter 9, only one of which deals with insolvency (Deal et al., 2013; Laughlin, 2005). (5) Federal bankruptcy code defines insolvency as a local government either "not paying its debts as they become due" or "unable to pay its debts as they become due" (Laughlin, 2005, p. 40). The first definition examines whether the government has paid past and current debts and the second definition examines on the government's inability to pay future debts. However, reports of case law suggest that the solvency criterion courts use focuses on prior and near-term financial obligations and the ability of governments to generate enough cash to pay these debts (Park, 2004).
For instance, Lewis (1994b) reports that the petition by the City of Bridgeport, Connecticut for bankruptcy in 1991, which was the largest ever general-purpose government petition at the time, was dismissed because the judge noted that "Bridgeport's insolvency should be judged by a cash flow, not a budget deficiency, analysis" (p. 520). From a different perspective, focusing on cash solvency can also mean a government that appears to have money could actually be judged as insolvent. In 2008, Vallejo filed for bankruptcy, and was sued by employee groups that argued the City was not insolvent due to the existence of reserve funds. These arguments were rejected, however, by the court due to restrictions on the funds that prevented the City from using them to pay for general operations (Scorsone & Bateson, 2012).
The court's emphasis on the ability to pay existing bills and debt service using available funds overlooks whether governments that file for bankruptcy have the long-term financial capacity to benefit from the tools that help such governments become solvent in the short-term and remain solvent in the future. The City of Detroit, the largest city in the US to file for bankruptcy in the US as of 2013, exemplifies a government that has little long-term financial capacity and fundamental financial problems that cannot be solved through bankruptcy. Restructuring its debt and union contracts and selling assets (including paintings in the Detroit Institute of Arts) will help the city to operate more cheaply and pay down its current liabilities (The Economist, 2013; Matthews, 2013).
However, short-term finances are not the true problem in Detroit. Detroit's population has declined from nearly 2 million in 1950 to just above 700,000 in 2010 (Borney & Gallagher, 2013). Of the remaining population, 60% live in poverty and 18% is unemployed. One third of Detroit's land area is considered to be vacant or derelict. Bankruptcy turnaround efforts will not insure that Detroit will be able to meet its spending obligations in the future and provide sufficiently for the health and safety of its population using only the declining revenue resources available within the jurisdiction.
2.3 Fiscal Crises: The Need for a Broader Focus.
The need to focus on long-term solvency to understand the reasons for bankruptcy and its effect are also apparent from other large cities that have experienced fiscal crises including Cleveland (1978), Philadelphia (1991), Newark (continuous), and Buffalo (2003). Similar to Detroit, these cities have entrenched fiscal deficits that developed over time due to reasons that are mostly beyond the control of elected officials in the local government. Suburbanization of the larger metropolitan regions and the decline of manufacturing jobs in the US have depleted these governments' revenue bases to the point where they cannot collect enough revenue to provide adequate services at a reasonable level of burden to the taxpayers remaining in the jurisdiction.
Additionally, these taxpayers often have greater spending needs that increase the financial obligations of the governments relative to other cities in the region or central cities in the U.S., all of which results in a continuing inability of these governments to pay their pills and meet debt payments. Although their financial problems at any one time may not meet the criteria of insolvency according to the bankruptcy standard, their financial condition is precarious enough in the long-term to keep them in a permanent state of poor fiscal health in the short-term. Without changes to their underlying financial capacity, these governments are destined to bounce from one fiscal crisis to the next, and the provisions afforded by bankruptcy are not likely to solve their long-term financial problems or ensure that they will not be candidates for bankruptcy in the future.
These governments stand in stark contrast with others in the U.S. such as Orange County (1994), San Diego (2004), Miami (1996), Jefferson County (2011), Vallejo (2008), and Boise County (2011) where fiscal crises seem out of place. Although some of these governments have filed for Chapter 9, they do not have the same structural imbalance between revenue bases and spending needs as the other group of governments. The income levels of residents in San Diego and Orange County are some of the highest in the nation, and their revenue bases are more than adequate to meet basic spending needs and citizen demands. Although less wealthy and with higher burden and spending needs, Vallejo, Miami, Jefferson County, and Boise County also do not have the structural deficits apparent in the first set of insolvent governments (Dluhy & Frank, 2002). Thus, bankruptcy in this second group of governments is probably a better tool for returning them to good financial condition compared to governments in the first group that have significant structural imbalances and less capacity to maintain solvency in the long run.
Another obvious factor in the fiscal crises in most governments in the second group is the occurrence of an adverse but uncommon event that resulted in spending obligations becoming much greater than the revenues and cash available to the government. In the case of Boise County, which is quite small, it was a lawsuit from a developer, and Vallejo's problems arose with the collapse of the housing market in California. San Diego's crises began with the discovery of an unreported massive pension deficit, and Orange County became insolvent when rising interest rates led to a significant loss of the value of invested funds (Baldassare, 1998). Jefferson County also experienced a massive loss of financial assets in the market due, in part, to the subprime mortgage meltdown.
It is also important to emphasize that, in most of these cases, the adverse event revealed numerous risky and poor financial decisions that made these governments vulnerable to a fiscal crisis, whatever the proximate cause. The crisis in Vallejo exposed over-generous labor agreements (pension and salary), and San Diego exposed illegal activities and sloppy financial reporting. The fund seizures in Orange County were the result of risky investments, and Jefferson County's losses were due to speculative bond transactions (Alabama Policy Institute, 2008).
Although adverse events and unsound or risky financial decisions are also likely to be significant contributors to the insolvency of governments with significant structural imbalances (Deal, Kamnikar, & Kamnikar, 2009), such governments are not likely to have enough reserve resources (revenues or cash) to compensate for the financial demands created by an adverse event or poor decisions. The revenue bases of these governments are often too poor and their revenue burdens too high to generate sufficient new revenue. Compared to wealthier governments, those with structural imbalances also are likely to operate with less cash reserves and a higher level of fixed costs making them less able to successfully manage fiscal crises (Hendrick, 2011). Additionally, these conditions will hinder the impact of financial management practices on overall financial condition of these governments and limit the range of fiscal policies officials can pursue to resolve fiscal crises and return the governments to sustainable solvency.
3. SOLVENCY, FINANCIAL CONDITION, AND RESEARCH QUESTIONS
The differences between these two sets of governments demonstrate that the concept of solvency or financial condition can be interpreted in several ways. More specifically, we can think of government solvency as occurring in more than one dimension or time frame. One of the most recognized approaches to framing the concept of solvency comes from Nollenberger et al. (2003) who distinguish between four types of solvency (See Hendrick, 2011, pp. 22-23):
* Cash solvency: the ability to generate enough cash over thirty or sixty days to pay bills during that time period. Measured with indicators such as liquidity ratio (cash + cash equivalents / current liabilities) and current ratio (current assets / current liabilities)
* Budgetary solvency : the ability to balance the budget and generate enough resources to cover expenditures over a normal budget cycle. Measured with indicators such as operating ratio (revenues / expenditures) and fund balance ratio (fund balance / expenditures)
* Service-level solvency: ability to provide adequate services to meet the health, safety, and welfare needs of its citizens given available resources. Measured with indicators of revenue burden (revenues per capita) and spending effort (spending per capita).
* Long-term solvency: the ability to balance revenues and spending, meet future obligations, and handle unknown financial challenges in the long run. Measured with ratios such as long-term liabilities / total assets and indicators of the wealth of the revenue base (e.g. income per capita and market value of homes) and spending needs (e.g. crime and poverty rate and pension obligations).
Using these categories of solvency we can restate our argument that bankruptcy case law in the US defines insolvency primarily as budgetary and cash solvency and not service-level and long-term solvency, which overlooks the fundamental and long-term financial problems of governments like Detroit, Newark, and Buffalo. Although governments that are insolvent at the service and long-term levels are more vulnerable to insolvency at the cash and budgetary levels due to adverse shocks, risky financial decisions, and poor financial management, insolvency in the short-term is less likely to drive insolvency in the long run and at the service level. Given the complex nature of the relationships between different levels of solvency, the financial condition of government with respect to bankruptcy should be judged in all areas of solvency. One might even argue that long run and service level solvencies are more important in the assessment of solvency due to the entrenched nature of these problems and the difficulty of resolving them with bankruptcy.
3.1 Financial Condition as a Balance among Attributes
Financial condition within each of these dimensions is best characterized as a balance among fiscal attributes within and across these dimensions and is measured using ratios of attributes (Clark and Ferguson, 1983). Insolvency, on the other hand, is best characterized as a discrete phenomenon in which the distinction between a solvent and insolvent government depends on whether it has reached a critical level of financial condition or imbalance among attributes beyond which it is insolvent (e.g. does not have enough cash to pay bills in 60 days or does not have enough uncollected revenue to meet minimum service levels).
More generally, defining financial condition using the concept of balance emphasizes that it is not specifically about the absolute level of resources available to government, their spending needs, or expenditures. Rather financial condition depends upon whether one set of fiscal attributes is appropriate to another as shown in Figure 1 (Hendrick, 2011, p. 21). In other words, it depends whether there are enough resources to meet spending needs (servicelevel solvency), whether revenues collected burden existing resources too much (budgetary solvency), whether expenditures are appropriate to revenues collected (budgetary solvency), and whether liabilities are balanced with assets (long-term solvency).
Defining financial condition as a balance of attributes within and across different dimensions of solvency and defining insolvency as a critical level of imbalance is generalizable to local governments throughout the US and other countries. For the most part, local governments in all countries collect revenue and make spending decisions that have impacts in both the long run and short run based on the level of fiscal resources available to them and the spending pressures they face. The U.S. is very different from other developed countries, however, in that local government officials in the U.S. control a much larger portion of the attributes that must be balanced. Most revenue received by U.S. local governments is "own source" (resources within the jurisdiction), but most revenue received by local governments in other developed countries comes from the national level of government. Additionally, the amount of own-source revenue collected by local governments in the U.S. varies by state. Thus, which government is responsible for a lack of balance among attributes of financial condition or insolvency at the local government level will vary by country and even U.S. state, but the concept of balance among attributes in different dimensions will apply irrespective of ownership of revenue resources.
This framework also provides a guide for organizing and capturing the complexity of financial condition and explaining fiscal crises, including bankruptcy and debt defaults, under different conditions. For instance, we expect that governments with poor long-term (and service-level) solvency are more likely to have poor short-term solvency (cash and budgetary) and that governments with good long-term solvency are likely to have good short-term solvency. We also know that there are governments with poor short-term solvency but good long-term solvency, such as San Diego and Orange County, and that poor financial management and risky fiscal policies are likely to have contributed to these governments' fiscal crises. We do not know, however, whether there are likely to be many local governments with good short-term solvency but bad long-term solvency. Although governments with poor long-term solvency are more exposed to adverse events than governments with good long-term solvency, we expect that governments with poor long-term solvency but good short-term solvency have benefited from good financial management and sound fiscal policy.
3.2 Using Solvencies to Predict Decisions that Create Imbalance
Table 1 shows how governments might be distributed on the continuums of long-term and short-term solvency. It also can be used to deduce the likelihood of fiscal crises in governments in some cells and the visibility of the contribution of financial management and financial decisions to the balance of attributes in other cells. Cell A represents governments like Detroit and Newark that are most likely to experience insolvency, and governments in cell D are least at risk of a fiscal crisis. Because short-term insolvency is often caused by long-term insolvency and the latter is often precipitated by events that are beyond the control of governments, it is impossible to know the role of officials' financial choices in creating the state of affairs of governments in cell A without a detailed investigation of events. Likewise, we cannot surmise the quality of financial decisions of officials in governments in cell D without further investigation because the attributes that contribute to good long-term solvency may mask or compensate for poor financial choices. Although poor financial decisions can create imbalances on both continuums, good long-term solvency allows governments to recover more easily from adverse fiscal events than poor long-term solvency.
On the other hand, we can deduce that governments in cell B are likely to be there because of good financial practices and policies that have allowed them to maintain good short-term solvency despite poor long-term solvency. We also can deduce that governments in cell C demonstrate the opposite of governments in cell B- governments with good long-term solvency can have poor short-term solvency and this is likely due to poor financial practices and policies. Governments such as San Diego, Miami, and Orange County would be classified into cell C.
3.3 Research Questions
Our analysis determines the long-term and short-term solvencies of suburban municipal governments in the Chicago metropolitan region during fiscal good times and bad times and then uses this classification system to answer the following questions:
1. How strong is the relationship between the different types of solvency, especially short-term and long-term solvency, for these governments and are the relationships linear?
2. What are the effects of fiscal good times and bad times on the solvencies of these governments and the relationship between long-term and short-term solvency?
3. To what extent do governments exist in cells B and C and what characteristics might explain their placement in these cells?
If the relationship between short-term and long-term solvency is strong and there are few governments in cells B and C, then this suggests that financial management and fiscal policy may have little effect on the overall solvency of governments. But, if the relationship is weak and there are numerous governments in these cells, then we can investigate these governments further to determine if there are common factors that might explain their solvency levels, including financial practices and decisions. The relationship between these two insolvencies also may be non-linear, such as if there are many governments in cell C but few in cell B. This outcome suggests the ease with which governments with good long-term solvency can weaken their short-term financial position and the difficulty of maintaining good short-term solvency under conditions of poor long-term solvency. This outcome also suggests that overall solvency cannot be easily characterized as a summation of component features or dimensions. The analysis also allows us to assess the effects of fiscal bad times, which is an adverse event, on these governments' solvencies and solvency features relative to fiscal good times.
4. CONTEXT AND RESEARCH DESIGN
4.1 Chicago, Illinois, and the United States in Context
According to the US Census Bureau, the Chicago metropolitan area includes 1,790 total local governments in 14 counties located in the three states. (6) More commonly, however, the region is recognized as consisting of six counties in northeastern Illinois. This area had a population of roughly 8.4 million in 2010 and 265 suburban, municipal governments (incorporated villages and cities) as of 2000. Cook County is the largest in the region with 63% of the regional population and 45% of the municipalities. The median population of municipalities in the region was approximately 12,500 in 2010, and the median total operating expenditures was about $13.5 million. One city in the region has a population of almost 200,000 (four cities have populations greater than 100,000), and the region has 20 villages with populations less than 1000.
Figure 2 shows a map of the municipalities in the Chicago region with the five largest jurisdictions in terms of population shown in heavy outline and labeled. Four of these jurisdictions (Waukegan, Joliet, Aurora, and Elgin) are recognized satellite cities, which had their own suburbs that predated the expansion of suburban development away from the City of Chicago.
Compared to other local governments in the U.S., municipalities tend to provide a broader range of local services than most counties and townships, and are considered to be general-purpose governments compared to special or single purpose school districts and special districts. Municipalities also have authority over land use and economic development within their boundaries. Thus, the solvency of municipalities is particularly important compared to other local governments in the U.S. with respect to their responsibilities in providing for the health and safety of citizens at the local level. In most states, welfare services are provided by state and county governments (these services are provided by townships and state government in Illinois), and primary and secondary education are provided by school districts. We should also mention that municipalities that are home rule in Illinois are not authorized to file for Chapter 9 bankruptcy (Chatz, 2010; Laughlin, 2005). (8)
Home rule status is automatically granted to municipalities in Illinois with populations greater than 25,000, but smaller municipalities can obtain home rule status via referendum and larger municipalities can also rescind it through referendum. (9) As of 2010, about 50% of the municipalities in the region were home rule, and over half of these were home rule via referendum with a population less than 25,000. Cook County is the only home rule county in Illinois.
Municipalities in Illinois also have the authority to levy a wide range of taxes, which also varies by whether the government is home rule. In addition to the property tax, municipalities in Illinois may levy taxes on cigarettes, photo finishing, motel occupancy, automobile rental, and utilities (telephone, natural gas, and electricity). Home rule municipalities can levy an even greater range of taxes, including an additional sales tax, and they are not subject to property tax limitations. (10) At the median in 2010, only about 15% of total revenue in the 265 Chicago municipalities comes from state government, which is a much lower percentage of intergovernmental aid compared to local governments in other developed nations. Of the remaining own-source revenue of governments in this study, about 40% (at the median) is from property tax, 20% is from sales tax, 13% is from other tax sources, and 22% is from charges and fees.
4.2 Research Design
Most of the financial data that we use to assess the solvency of these governments and conduct our analyses comes from published financial information from the Illinois Office of the Comptroller (IOC). This agency requires all local governments in the state to submit information from the yearly Annual Financial Reports (AFRs), which are externally audited. Other financial data were obtained from the Illinois Department of Revenue and all financial data were obtained for 1997 to 2010. Socio-economic and demographic data came from the US Bureau of the Census (decennial censuses for 1990, 2000, and 2010 and the American Community Survey for 2010, 5-year estimates). Most of the financial indicators used to measure solvency and trends in finances are calculated as three-year moving averages for the entire time period of data, which yields estimates for 1998 to 2009. (11)
To distinguish between fiscal good years (munificence) and fiscal bad years (stress) we examined percent change of values across all years of data for nine indicators that represent trends in the underlying revenue bases and other factors in the solvency of these governments. These indicators are enumerated and explained in Appendix A. The median percent changes are used for all nine indicators because the skewness of the distributions makes the means less useful as a measure of central tendency. The medians of all indicators combined suggest that the years 1998, 1999, 2000, 2001, 2005, and 2006 could be classified as fiscal good times, and the years 2002, 2003, 2004, 2007, 2008, 2009 could be classified as fiscal bad times.
Table 2 shows these percent changes for three of the indicators--percent fund balance, days of cash on hand, and total revenues (corrected for inflation). If the medians are summed for the three indicators in table 2, the sums for the good years are positive and the sums for the bad years are negative, with the exception of 2007, which has a sum of 3. However, looking at change in sales receipts, change deficit ratio, and change in expenditure coverage (not shown here) the year 2007 shows declines in all these trends and so was designated as a bad year. Similarly, 2001 could be designated a bad year with a sum of only 1.2 for all the medians in Table 2, but other indicators showed increases in trends and so 2001 was designated as a good fiscal year.
Appendix A and the next section also describe the five solvency measures that we assess with correlations, scatterplot graphs, and cross-tabulation tables to answer our three research questions. Four of these indicators are aggregated in the analyses based on whether the years of observation are classified as good times or bad times. We also use cross-tabulation tables to identify municipalities in the cells of Table 1 that are at the extreme of the distribution for each pair of variables and then examine these municipalities for common causal factors. This approach represents a macro-comparative methodology that utilizes values on the dependent variable, which are solvency measures in this case, and outliers to generate samples of municipalities. Following Mill's method of agreement, the governments in these samples are then examined for common characteristics that may explain their solvency, which can then be tested in other populations of governments (Bollen, 1993).
5. MEASUREMENT OF SOLVENCY AND METHODOLOGY
Our analysis employs the following five measures of solvency that are described in detail in Appendix A. 1) revenue wealth / spending needs (longterm solvency); 2) expenditure coverage (short-term solvency); 3) days of cash on hand (short-term solvency); 4) revenue burden (service-level solvency); and 5) general obligation (GO) debt burden. The last indicator in this list might be considered a factor in long-term solvency because it affects longterm liabilities and claims on future revenues. However, unlike revenue wealth and spending needs, debt burden is directly controlled by the government and so it is analyzed separately.
5.1 Long-Term Solvency.
The index developed here to measure long-term solvency is calculated using an operationalization developed by Hendrick (2004, 2011) that focuses on local governments' revenue-raising capacity and spending responsibilities (Ladd & Yinger, 1989). Both revenue-raising capacity and spending responsibilities are, for the most part, exogenous factors in the model of local government financial condition that determine the level of revenues available to governments to meet the service needs of citizens (Hendrick, 2011, figure 2.2). Similar to Ladd and Yinger, we construct separate indicators of revenue wealth and spending needs and then look at their ratio to determine long-term fiscal health. Unlike Ladd and Yinger, however, our revenue wealth and spending needs indices are calculated using an estimate of the impact of components of these constructs on revenues and spending rather than measuring how individual governments compare on all components to the larger group (Sohl, Wood, Peddle, Kuhn, & Thurmaier, 2009). (12)
As described in the prior section, the primary revenue bases for municipalities in Illinois are EAV, sales receipts, and personal income. A prior financial study also shows that high population growth, which is common in outlying jurisdictions, has a significant effect on the spending and revenue flows of municipal governments in the region (Hendrick, 2011). Because home rule privileges greatly affect these governments' access to revenue sources, this feature is also recognized in assessing revenue wealth. With respect to spending needs and costs, this index is constructed by measuring the following factors (and variables): the impact of aging infrastructure (median age of housing), crime levels (public safety spending), population density (costs per capita), residential land use (responsibility for residential services), miles from the central city (services per capita), and whether the government is in a fire district (services per capita). (13)
Based on the argument that spending needs and revenue wealth are exogenous, less controllable by the government, and relatively entrenched, we calculate only one index of long-term solvency for the entire period of the study (1998 - 2009) rather than separate indices for each year of data. Additionally, data for median age of housing and personal income is available only for 2000 and 2009 and population figures are most accurate for 2000 and 2010. Thus, the revenue wealth and spending needs indices are calculated using pooled data for 2000 and 2009 creating only one value for long-term solvency for each government.
As shown in Appendix A the revenue wealth variables are regressed against own-source revenue per capita and the spending needs variables are regressed against operating expenditures per capita. The standardized slopes of these models are multiplied by the percent medians for all variables in the model for each government in the region and then summed to create two positive indices. (14) The index for long-term solvency is then calculated as revenue wealth / spending needs and reflects the balance between these two fiscal characteristics.
Unfortunately, the values of this index are not readily interpretable as a percentage or recognizable function, but the index does serve the purpose of measuring the relationship between these two features to assess the government's long-term solvency. The predictive validity of this method of measuring revenue wealth and spending needs was tested extensively in Hendrick (2011 and 2004), and it is also tested here by examining correlations between these indices and other financial indices available in the larger financial database, including the other solvency measures. (15) These correlations showed that the long-term solvency measures have good predictive validity with other financial indicators, including the solvency measures examined here. (16)
Appendix B shows descriptive statistics for the variables used to construct the indices and other variables of interest. This appendix shows the tremendous variation in financial attributes and other features of the municipalities in this region. For instance, the smallest city in 2009 had a population of only 89 and the largest city had a population of over 185,000. In terms of personal income per capita of citizens in these jurisdictions, the figures range from $8,510 to $103,943, which shows tremendous inequality in the wealth of citizens served by the municipal governments in the region. Similarly, the equalized assessed value per square mile of the jurisdictions, which is the base for property taxes, ranges from $5,713 to over $6.33 million. With respect to own-source revenues and operational spending per capita, these governments range from $37 to $7,696 and $8.70 to $3,180 respectively. Clearly, some governments in this region collect very little revenue and spend very little on their citizens compared to other governments that collect a lot of revenue and spend a lot.
Figure 2 shows how municipalities in the Chicago region are distributed on the wealth-need index. This pattern is consistent with common understanding of the basic wealth and spending needs of populations and governments within the region. The figure shows financially depressed and needy areas just west of the City of Chicago and throughout the southern part of Cook County. These governments were part of the first stage of migration of population, industry, and commerce from the central city after World War I and World War II. Financially depressed and needy areas also occur around the four satellite cities that are part of the third ring of suburbs at the edge of the developed area of the region. The wealthiest areas relative to spending needs are just north of the central city and distributed throughout middle ring suburbs, which grew primarily during the second wave of migration from the central city in the 1960s and 1970s (Conzen, 2005; Keating, 2005).
5.2 Other Solvency Measures.
With respect to revenue burden, a tax rate is a measure of the burden of one tax collected by a government on the tax base for that revenue source (also called the tax effort of a government), but it does not make sense to sum tax rates or charges to get an aggregate measure of overall revenue burden. To obtain such an index, Ladd and Yinger (1989) divided total revenues collected by their measure of revenue capacity. We follow suit by dividing total own source revenues collected by each municipality by its measure of revenue wealth. Others have used revenue per capita as a measure of overall revenue burden on the residents of a jurisdiction and service-level solvency (Wang, Dennis, & Tu, 2007), but this indicator does not reflect the burden of revenue collection on all revenue bases.
The expenditure coverage index merges the operating ratio (Wang, et al., 2007) with a standardized measure of government fund balance (Maher, 2013). The operating ratio (revenues / operating expenditures), which shows the extent to which revenues covers spending, is also a measure of the level of deficit or surplus at the end of the fiscal year. Fund balance, which is assets minus liabilities, as a percentage of operating expenditures shows the extent to which surplus revenues covers spending. (17) Decisions about fund balances and operating deficits and surpluses at the beginning of the fiscal year and during budget execution are often made in conjunction with each other because one directly affects the other in the accounting process. Thus, we combine the two measures to determine what percent of governments' expenditures are covered by both current and surplus revenues. The second short-term solvency measure, days of cash on hand ((cash + investments) / (operating expenditures / 360)) is a widely recognized measure of liquidity in the financial field.
We measured debt burden in these municipalities as total GO debt divided by EAV because in the State of Illinois the GO debt of local governments is secured only by the property tax and not the entire taxing power of the government. Although municipalities in the state may also issue revenue bonds to finance capital projects for services, such as water and sewer that are funded entirely by charges, most of them issue GO bonds because the interest rates are lower. These governments can also issue alternative revenue bonds that are secured by other taxes, but these bonds are even less prevalent than revenue bonds among these governments.
To answer our first two research questions, our analysis focuses on examining the bivariate relationships between all five solvency indices but especially between the indices that represent short-term and long-term solvency. For the four indices that are calculated separately for each
year of data (#2 to #5), we calculate the mean for the good times and bad times separately to obtain an aggregate measure for each index for these two time periods. We examine correlations and scatterplots of the relationships between these aggregated indices, and transform them into septiles with an equal number of cases to examine cross-tabulation tables for pairs of variables to answer the third research question. We then focus in on the municipalities that are in the corners of the short-term and long-term solvency tables and examine what features these governments share, especially governments in cells B and C in Table 1, to determine if their position might be explained by features such as size, growth, or whether their financial management practices and fiscal policies should be examined further. We also examine how governments in these cells and especially cell A are distributed on GO debt burden and revenue burden (service-level solvency) in order to identify and predict the causes of insolvency.
The first two research questions are answered in the next section and the last research question is answered in the section following the next one. Although we do not investigate the financial management practices and fiscal policies of these governments for this study, we use reports from Hendrick (2011) to provide information about many of governments identified in the cells of Table 1. These reports are based on interviews with the chief financial officer in 62 governments in 2003 and extensive analysis of news reports on a broader set of governments in this region from 2000 to 2006.
6.1 Relationships between the Solvency Measures.
Examining the scatterplots and the correlations between the four aggregated indices for fiscal good times and bad times shows that these indices are highly correlated for these two time periods as demonstrated in Table 3. The index days cash on hand for good times and bad times is the lowest correlation at .82 and revenue burden is the highest at .96. We used the scatterplots to identify extreme outliers in each relationship that might inflate the correlations and removed these cases prior to calculating the correlations. We also removed 2004 and 2007 from the aggregated indices for fiscal bad times and 2001 for fiscal good times and recalculated the correlations to determine if these years, which are not clearly good times or bad times (see Table 2), affected the relationships. These new correlations are only slightly greater than those in Table 3.
Both sets of correlations show that the conditions measured by these indices do not change across municipalities for fiscal good times and bad times. Although the trend analysis of these indices and other financial characteristics show that solvency conditions change within these governments during the different periods, the relative conditions of these governments do not change. In other words, compared to each other, governments that have high revenue burden in good times will also have high revenue burden in bad times and vice versa. Thus, we can infer that relative solvency is fairly consistent among these governments for the different time periods. On the other hand, the relationships between the five different solvency measures paint a picture of diversified conditions in these governments.
Table 4 shows the correlations between the five solvency measures for fiscal good times and with extreme outliers removed. Notice the very high correlation between expenditure coverage and days of cash on hand, which is due, in part, to the fact that both measures of short-term solvency are divided by operating expenditures. The weakest correlations exist between wealth relative to spending needs (long-term solvency) and the two short-term solvency measures. Although the relationships are consistent with the expectation that governments with high long-term solvency will tend to have high short-term solvency and vice versa, the scatterplots show that there are a many governments with high long-term solvency and low short-term solvency (cell B in Table 1). There are also many governments in cell C that have low long-term solvency but high short-term solvency. Also consistent with expectations, governments with high wealth relative to needs have lower revenue burden and lower GO debt burden. High GO debt is also associated with lower shortterm solvency and higher revenue burden, and higher revenue burden is associated with weaker short-term solvency.
It should also be noted that all correlations in Table 4 are somewhat lower for the indices aggregated for fiscal bad times even after outliers are removed, which shows that the relationships between the different areas of solvency hold throughout the entire time period. Additionally, scatterplots for almost all negative relationships show that these relationships are significantly convex to the origin (similar to an indifference curve), which indicates that these relationships are not consistent across the range of values of the indices. With respect to cash solvency and revenue burden, for instance, the pattern in the scatterplot shows that cash solvency increases faster among governments as their revenue burden declines and revenue burden increases faster among governments as their cash solvency becomes lower.
Overall the correlations between the solvency measures and the scatterplots demonstrate the complexity of the relationships between the different areas of solvency and that municipal governments in the Chicago region are distributed broadly across the joint distributions of these indices. With respect to Table 1, we also know that there are likely to be many governments in cells B and C. The next section examines governments in these cells in more depth and also examines governments that have a low level of solvency on all indices (cell A).
6.2 Analysis of Solvency in the Extreme.
Dividing all solvency measures for both good times and bad times into septiles with an equal number of cases in each group, we identify and examine the municipalities that are in the lowest two septiles of both measures of short-term solvency and the lowest two septiles for long-term solvency (for either good times or bad times). This process yields 43 governments in the region in which financial crises are likely to occur (cell A Table 1). Table 5 shows a crosstabulation using septiles of revenue wealth relative to spending needs (long-term solvency) and spending coverage (short-term solvency) for fiscal good times with the municipalities in cell A designated in gray in the upper left-hand corner of the table. Governments with high wealth relative to need and low spending coverage (cell C) are shown in gray in the upper right-hand corner of the table, and governments in cell B with low long-term solvency and high short-term solvency are shown in gray in the lower left-hand corner of the table. (18)
Looking more closely at the governments in cell A, we observe that they have a wide range in population from 1,049 to nearly 90,000 residents. However, most of them are in South Cook County or just west of the City of Chicago, and many were previously identified by Hendrick (2011) as having significant financial problems, political conflict, and weaknesses in financial management practices. Eleven of these governments also have high revenue burden and GO debt burden (at the 5th septile or above), and so are the worst off in this group in terms of overall solvency and risk of financial crises. Interestingly, these governments all have populations less than 20,000 and only two of them are home rule. Nineteen of the 43 governments in cell A have high revenue burden but moderate debt burden, and most of these are home rule. This finding suggests that home rule is important in helping municipal governments in this region avoid the lowest level of solvency on all five measures.
Five of the 43 governments in cell A have both low revenue and low debt burden, which suggests they may follow a policy of 'penny pinching.' Although these governments are at some risk of fiscal crises, they may purposely keep low levels of cash and surplus funds in conjunction with low debt, taxes, and charges. These governments seem to be living within their means compared to other governments in this cell and purposely balancing the controllable features of their financial condition with the uncontrollable features of their fiscal environment. Such policies were confirmed for three of the governments in this group by Hendrick (2011), who also noted that these three governments had significant underfunding of capital infrastructure.
With respect to low long-term solvency and high short-term solvency, the question is to what extent do governments exist in cell B of Table 1 and what explains their level of financial condition? Our analysis shows only 21 governments in this category. Only three of these governments have populations greater than 6,000, only four are home rule, most have experienced population growth (median and mean growth rate of 20% and 119% respectively), and most are located at the edges of the developed area of the region that are serviced by fire districts. Twelve of these governments, which include the three governments in the group with populations greater than 13,000, have high revenue burden and half of these also have high debt burden. Given the small size, location, rural nature, and growth of most of these governments, especially those with low debt and revenue burden, it would be hard to attribute their good short-term solvency to good financial management. Rather, their solvency features, including high debt and revenue burden, are more likely a function of their high growth. The three larger governments of this group may be the exception here, and deserve further investigation to determine whether good fiscal policy and practice helps these governments to maintain good short-term solvency despite have poor long-term. In this case, Hendrick (2011) did identify the largest government with a population of 27,000 as having good financial management practices (and low debt).
Compared to the governments in cell B, some of those with very high long-term solvency and very low short-term solvency are likely to be at risk for a fiscal crisis in the same manner as governments such as San Diego, Orange County, and Vallejo. Our analysis identified 23 governments in this group, 13 of which have low debt and low revenue burden and so are not as fiscally leveraged as the other ten governments with moderate to high debt and revenue burden. Governments in cell C range in population from about 4,000 to nearly 57,000 residents with the mean and median population about 21,000. Most of these jurisdictions are in the middle ring suburbs that are north and west of the City of Chicago, and they had either slow growing or declining populations during the time period of the study (median population growth of 1%). Almost half the governments are home rule and a little over half of them have fire services provided by a separate fire district.
Although the likelihood of a fiscal crisis happening in governments in cell C is probably remote, it would be worthwhile to investigate the extent to which their fiscal policies and practices have contributed to the poor short-term solvency in these governments, especially the ten governments with moderate to high debt and revenue burden. Hendrick (2011) conducted indepth investigation of two of these 10 governments and found them to have good financial management practices, but others in this group and in the larger group in cell C have reputations in the local news for risky financial policies, strong politics, and financial management problems more generally.
7. DISCUSSION AND CONCLUSION
One main purpose of our study is to examine the relationship between long-term and short-term solvency within a population of local governments to determine how they are jointly distributed on these features and then consider what this reveals about the characteristics of these governments and their insolvency more generally. We find that the relationship between short-term and long-term solvency is weak among municipal governments in the Chicago metropolitan area, which shows that these dimension do not always coincide and, when considered separately, may not predict a fiscal crisis. The number of governments in our study with good long-term solvency and poor shortterm solvency (cell C Table 1) shows that this state of affairs is not uncommon. It also shows that a high level of fiscal comfort and opportunity does not guarantee that a government can pay its bills or meet other financial obligations. On the other end of the continuum, governments with very poor longterm solvency but good short-term solvency (cell B) also exist in this region. Compared to governments in cell C, those in cell B seem better able to maintain operations and good financial position despite entrenched fiscal constraints and service demands.
Closer examination of some governments in cell B, however, shows that their placement in this cell may be due to their small size, location in the region, and population growth. If these environmental factors rather than effective financial management and sound fiscal policies explain these governments' good short-term solvency, then we may conclude that overcoming poor long-term solvency with policy and practice alone is a difficult task. On the other hand, the concentration of governments in cell C that cannot be explained by environmental factors indicates how easy it is for relatively wealthy governments to weaken their solvency through bad financial management and unsound fiscal policies. Similar to governments like Orange County and San Diego, some of the governments in cell C in the Chicago region may be one adverse event away from a major fiscal crisis or insolvency.
Another important finding is that this region has a relatively high concentration of municipal governments with both poor short-term and long-term solvency (cell A) that also have high debt and revenue burden. This finding suggests that, at some point, it may be difficult for governments with very poor financial condition on all dimensions and levels of solvency to improve their fiscal health. Such governments are similar to central cities in the US that we normally associate with insolvency and extreme fiscal crises such as Detroit, Newark, and Cleveland. The potentially insolvent suburban municipalities identified here, however, may be worse off than these central cities because the suburban governments do not have the same level of leverageable assets. Thus, it might be more difficult for suburban and small governments to help themselves compared to central cities in a similar financial position.
These findings provide valuable lessons for conducting future research on the causes of insolvency in local governments and solutions to this problem. Specifically, our research shows that the mechanisms that drive poor short-term solvency are likely to be different than the mechanisms that drive poor long-term solvency. Our findings also demonstrate that focusing analysis of the causes of poor financial condition and solutions to this problem on governments in cells C and B could be more productive than targeting the broader population of governments or governments in cell A. It is not as easy to separate out the causes of poor short-term solvency of governments in cell A as governments in cell C. Similarly, it is easier to see the impacts of officials' efforts to improve short-term solvency among government in cell B compared to governments that have a moderate level of long-term solvency.
The joint distributions of municipal governments in this region on the different dimensions of solvency also demonstrate that assessing their financial condition and predicting fiscal crises may not be as simple as summing together a set of indicators across different dimensions as has been the practice in prior research (See for example Brown, 1993, Wang et al., 2007, and Kloha, Weissert, & Kleine, 2005). Our analysis has shown that the solvency dimensions of the governments that are examined here are related in complex and non-linear ways that may require a higher weighting of some dimensions or components of dimensions compared to other dimensions and component to obtain a valid measure of overall financial condition in governments in the US and elsewhere.
The high concentration of local governments in this region in cell A begs the question of whether any of these governments might actually be insolvent and incapable of meeting their financial obligations, including the obligation to provide adequate services in the areas for which these governments are responsible. To our knowledge, none of the governments in the Chicago region with the authority to file for Chapter 9 bankruptcy have ever done so, and none have ever asked the State of Illinois for financial assistance under laws passed in 1988 and 1990. 19 There is, however, evidence that the public safety has been threatened by the poor finances of several governments in cell A of Table 1. Numerous rapes that occurred over a period of 20 years were not investigated in one financially "struggling" government in cell A, and police services in two other governments in cell A were taken over by the county government because the municipal governments could no longer afford to provide this service (Sweeney, 2013).
This finding raises another question of why there is little action on the part of Illinois to change the financial condition of potentially insolvent local governments, and whether this trend exists in other states. As noted previously, there are strong incentives for all parties to avoid bankruptcy by local governments. Similarly, there may be strong incentives for both the state and local officials in the US to ignore local government insolvency. Current economic conditions at all levels of government and the system of state and local relationships in the US may give elected local officials little incentive to relinquish power to their state government. These circumstances may also give state government officials little incentive to assume responsibility for insolvent local governments (Nowlan, Gove, & Winkel, 2010). Thus, many of the governments in cells A may be destined to limp along forever from one fiscal crisis to the next.
Concerning the effects of fiscal good times and bad times of the governments in this region, we found that most fiscal solvency indicators were lower during fiscal bad times compared to fiscal good times, but this did not change the relative solvencies of these governments across the septile groups. Rather, most municipalities remained in the same septile group for fiscal good times and bad times. This finding indicates that recessions are not likely to rearrange the solvency order of governments and suggests that recessions or other causes of fiscal stress are not likely to result in governments becoming insolvent or to create a fiscal crisis in governments that had good financial condition prior to the event. However, we did find that the relationships between the five areas of fiscal solvency during fiscal good times that are shown in Table 4 were all weaker for fiscal bad times, which indicates that adverse fiscal events can have somewhat different effects on individual governments.
Although our research does not present a comprehensive analysis of the causes of fiscal insolvency in local governments or suggest strategies for how such governments can improve their financial condition, it does provide tools and a plan of analysis for future research. Specifically, it provides a way of conceptualizing and observing solvency in local governments in order to identify it and determine its distribution in a population of governments. This method of classifying governments according to dimensions of solvency yields important clues about broad causes of insolvency in local governments with respect to the balance among these dimensions and components of the dimensions. Going forward, this system of classification provides a basis for investigating particular causes of insolvency among a population of local governments, and it establishes hypotheses about the relationship among the dimensions that could be tested in other states and countries. More importantly, our study suggests the extent to which long-term solvency is entrenched and difficult to alleviate and that solutions to insolvency must address long-term aspects of financial condition in order to be successful.
The relevancy of these hypotheses for local governments in other nations depends greatly on the broader governing and fiscal contexts of these governments, especially the fiscal authority and responsibility of higher versus lower levels of government. Whether these hypotheses are applicable to other nations also depends on the extent to which insolvent local governments may even exist in other nations. Our analysis has made clear that the financial fate of local governments in the U.S. depends greatly on the resources owned by these governments and that state government, which is the parent of local governments in that nation, may claim little responsibility for the solvency of lo cal governments. In other developed countries the national government is usually the parent of local governments and has a pattern of helping these governments to a much greater degree than state governments in the US. Thus, long-term solvency may be much more important to the financial health of local governments in the U.S. compared to other developed countries where the high level of aid to local governments insures that they meet some minimum level of long-run solvency. Moreover, factors affecting short-term solvency will be more important to the financial health of local governments in other developed countries than local governments in the U.S.
Most local governments in the U.S. use a funding accounting system that divides and maintains all transactions in three basic fund types: governmental, proprietary, and fiduciary. Governmental funds are further subdivided into the general fund (to account for all regular operations), the debt service fund (to account for principal and interest to repay all debt), multiple capital funds (to account for all spending and revenue devoted to infrastructure and equipment, and multiple special service funds (to account for earmarked revenues and spending and grants). Proprietary funds are further subdivided into multiple enterprise funds (to account for operations that are run like a business, such as sewer and water) and internal service funds (to account for services provided by one governmental unit to another).
Fiduciary funds account for assets held by government for someone or something else, such as pensions or taxes collected for other governments. Proprietary and fiduciary funds use a full accrual basis of accounting and governmental funds use a modified accrual bases. A full accrual system reports whether a fund is better or worse off economically and recognizes revenues when they are earned and expenses when they are incurred, regardless of when the cash is received or paid out. A modified accrual focuses more on the flow of resources within a fund rather than its economic value. It recognizes revenues when they are available and measurable and expenditures if they are to be made in the current fiscal year or soon thereafter.
Even within the modified accrual system there is a lot of variation within and between governments in the fiscal year in which transactions are recorded. This is the main reason that all financial variables are calculated as three year moving averages.
Unless indicated otherwise, all financial variables represent totals for all governmental operating funds: general, special revenue, and debt service (capital funds are excluded except where indicated). Governmental operating funds do not include enterprise funds (which are reserved primarily for water and sewer operations). Capital funds are for all purposes (governmental and enterprise). Percent change indicators used to determine fiscal good times and bad times (using three-year moving averages) are shown below.
1. Equalized assessed value (EAV): property tax base
2. Sales receipts: sales tax base
3. Intergovernmental revenue: shared revenue and grants that come primarily from state government
4. Days of cash on hand = (cash + investments) / (operating expenditures / 360)
5. Deficit ratio = (revenues / operating expenditures) *100
6. Percent fund balance = (assets - liabilities) / operating expenditures
7. Expenditure coverage ratio = ((revenue + fund balance) / operating expenditures) * 100
8. Total revenues (corrected for inflation)
9. Total operating expenditures (corrected for inflation)
Fiscal Good Times: 1998, 1999, 2000, 2001, 2005, and 2006
Fiscal Bad Times: 2002, 2003, 2004, 2007, 2008, and 2009
FINANCIAL CONDITION / SOLVENCY INDICES
All per capita measures except personal income per capita are weighted by multiplying them by percent residential EAV to correct for the distortion of per capita measures in non-residential municipalities. All financial indicators of solvency and components of the indicators (except some of those for long-term solvency) are calculated using three-year moving averages. All variables except percentages are corrected for inflation (ECI, 2000 = 100).
1) Long-Term Solvency: Revenue Wealth / Spending Needs
The indices revenue wealth and spending needs are estimated using pooled data 2000 and 2009 to create one measure of long-term solvency that applies for all years of data in the analysis (1998 - 2009). Revenue wealth measures the capacity of the government to generate own-source revenues from the main revenue bases, which are EAV, sales receipts, and personal income per capita. Home rule status and the ratio of population change are also used to estimate these indices. Each index is estimated separately using regression analysis as indicated below to obtain the standardized regression slopes, which are then multiplied by the percent medians of the regression variables and summed together to create an index.
Revenue Wealth: Regression analysis is estimated with the following variables. Own-source revenue is comprised of property tax, sales tax, other taxes, and charges and fees.
Dependent variable: Own-source revenue per capita (weighted)
A) Per capita personal income:
B) Sales receipts per capita (weighted):
C) Equalized assessed value per square mile:
D) [Population.sub.t] / [Population.sub.t-1] (logged)
E) Home rule status (0,1)
F) Year- 2009 (0,1); this dummy variable is included to reflect the fact that the index is calculated with data that are pooled for two years - 2000 and 2009
Revenue Wealth = .408A + .376B + .074C + -.037D +.17E + .158F; [R.sup.2] = .45
Variables A through D are recalculated as percent medians before multiplying with slopes to create index
Spending Needs: Regression analysis is estimated with the following variables. Some of the variables are reversed as indicated with an R. The purpose of this change, which does not change the relative values of the index, is to insure that all component variables in the index move in the same direction and that the index will be positive. Keeping negatively correlated component variables in the original direction creates some negative index values, which cannot be used accurately in the Revenue Wealth / Spending Needs ratio to reflect higher or lower levels of wealth relative to need.
Dependent variable : Operating expenditures per capita (weighted)
G) Median age housing (R):
H) Crime per 1000 population (weighted): These values are estimated for the 25 municipalities that report no crime statistics using expected values calculated from regression analyses of governments with reported crime figures.
I) Population density (R):
J) Miles from central city (R):
K) Percent residential EAV
L) Whether in fire district (R, 0,1):
M) Year- 2009 (0, 1) this dummy variable is included to reflect the fact that the index is calculated with data that are pooled for two years- 2000 and 2009
Spending Needs = .260G + .059H + .325I + .339J +.174K + .230L + .394.M [R.sup.2] = .42
Variables G through K are recalculated as percent medians before multiplying with slopes to create index
Short-Term Solvency: Two indices are used to estimate this level of solvency. Both indices are calculated for all years and then the mean index is calculated for good times and bad times separately.
2) Expenditure Coverage (budgetary solvency): (revenue + fund balance) / operating expenditures
3 Days Cash on Hand (cash solvency): (cash + investments) / (expenditures / 360)
4) Revenue Burden (service-level solvency): Own-source revenues / revenue wealth
5) General Obligation Debt Burden: General obligation debt / $1,000 EAV
APPENDIX B DESCRIPTIVE STATISTICS OF INDICES AND OTHER VARIABLES Equalized Per capita Sales assessed Ratio personal receipts per value per population income capita square mile change Mean $32,359 $10,615 $870,263 134 Median $27,469 $8,153 $661,618 108 Std. Deviation $15,958 $10,788 $757,912 185 Minimum $8,510 $37 $5,713 71 Maximum $103,943 $112,266 $6,333,928 3,655 Median Crime per Home Rule, build year of 1000 2003 housing population Mean Non-home 1972 33.0 Median rule, N = 160, 1973 23.5 Std. Deviation 60% 14.3 56.4 Minimum Home rule, 1939 .78 Maximum N = 105, 40% 2002 966 Miles from % Population City of Residential density Chicago EAV Mean 3,160 30.3 77.5 Median 2,739 28.0 81.3 Std. Deviation 2,438 13.5 16.3 Minimum 17.5 8.0 5.4 Maximum 15,378 68 100 Expenditure Whether in Population coverage fire district 2009 (good times) Mean Not in fire 18,397 214 Median district, N = 11,810 187 Std. Deviation 135, 51% 21,759 95 Minimum In fire district, 89 64 Maximum N =129, 186,991 671 General Days of cash obligation Own-source on hand debt burden revenues per (good times) (good times) capita Mean 379 2.65 $649 Median 280 1.39 $504 Std. Deviation 538 4.43 $654 Minimum 38 0.00 $37 Maximum 7,846 54 $7,694 Operational spending Deficit ratio % Fund bal- per capita (2005) ance (2005) Mean $436 121 51 Median $364 113 50 Std. Deviation $300 52 275 Minimum $8.70 68 -122 Maximum $3,180 753 539 Intergov't Revenue Revenue revenue per Wealth Burden Spending capita (2005) Index (good times) Needs Index Mean $16.9 103 1.31 127 Median $15.3 90 1.02 125 Std. Deviation $8.6 60 1.12 23 Minimum $0.63 17.8 0.28 98 Maximum $61 497 11.4 452 Revenue Wealth / Spending Mean 0.81 Median 0.74 Std. Deviation 0.43 Minimum 0.12 Maximum 3.47
Advisory Commission on Intergovernmental Relations (1973). City Financial Emergencies: The Intergovernmental Dimension. Retrieved April 14, 2013, from http://www.library.unt.edu/gpo/acir/Reports/policy/a-42.pdf
Advisory Commission on Intergovernmental Relations (1985). Bankruptcies, Defaults, and Other Local Government Financial Emergencies. Retrieved April 14, 2013, from http://www.library.unt.edu/gpo/acir/Reports/policy/a99.pdf
Alabama Policy Institute (2008, March 14). Jefferson County on Verge of Making Bankruptcy History. Retrieved from http://www.alabamapolicy.org/gary blog/article.php?id art=290
Baldassare, M. (1998). When Government Fails: The Orange County Bankruptcy, Berkeley, CA: University of California Press.
Berne, R. & Schramm, R. (1986). The Financial Analysis of Governments. NJ: Prentice-Hall.
Bollen, K. A., Entwisle, B. & Alderson, A.S. (1993). "Macrocomparative Research Methods," Annual Review of Sociology, 19: 321-351.
Borney, N., & Gallagher, J. (2013, September 15). How Detroit went broke: The answers may surprise you - and don't blame Coleman Young [Electronic version]. The Detroit Free Press.
Brown, K. W. 1993. The 10-Point Test of Financial Condition: Toward an Easy-to-Use Assessment Tool for Smaller Governments. Government Finance Review: 9(1), 21-25.
Chatz, J. A. (2010, September 20). Financial Reform for Municipalities in Illinois. In Illinois Municipal League. Retrieved November 9, 2011, from http://www.iml.org/page.cfm?key=5621 (Archived by WebCite[R] at http://www.webcitation.org/634nz7qZG)
Clark, T.N. & Ferguson, L.C. (1983). City Money: Political Processes, Fiscal Strain and Retrenchment. New York: Columbia University Press.
Conzen, M. P. (2005). Growth of the Chicago Metropolitan Area. In Encyclopedia of Chicago. Retrieved May 24, 2013, from http://www.encyclopedia.chicagohistory.org/pages/1756.html
Deal, K. H., Heier, J. R., & Kamnikar, J. A. (2013, Spring). 40 Years Later: An Analysis of Current Municipal Bankruptcy Cases. The Journal of Government Financial Management, 62(1), 26-32.
Deal, K., Kamnikar, J., & Kamnikar, E. (2009, Fall). A Descriptive Case Study of the Greene County, Alabama Bankruptcy. Journal of Public Budgeting, Accounting, & Financial Management, 21(3), 337-365.
Dubrow, D.L. (2009). The Treatment of Municipal Debt Under Chapter 9 of the Bankruptcy Code. Washington, DC: Arent Fox: http://www.arentfox.com/email/dubrow/Chapter%209%20Article.pdf
Dluhy, M. J. & Frank, H.A. (2002). The Miami Fiscal Crisis: Can a Poor City Regain Prosperity? Westport, CT: Praeger.
Driscoll, S. F. (2011, November 14). Washington Park bankruptcy case tossed. In Rockford Register-Star. Retrieved December 14, 2011, from http://www.rrstar.com/blogs/seanfdriscoll/x865539375/Washington-Parkbankruptcy-case-tossed (Archived by WebCite[R] at http://www.webcitation.org/63wFFBie2)
The Economist (2013, July 27). Detroit's bankruptcy: Can Motown be mended? Retrieved October 27, 2013, from http://www.economist.com/news/united-states/21582285-americas-biggestever-city-bankruptcy-starts-roll-can-motown-be- mended
Farmer, L. (2013). "The 'B' Word: Is the Stigma of Municipal Bankruptcy Going Away?" Governing Magazine, March, 36-41.
Gillette, C. P. (2012, January). Fiscal Federalism, Political Will, and Strategic Use of Municipal Bankruptcy. The University of Chicago Law Review, 79(1), 283-331.
Hendrick, R.M. (2004). "Assessing and Measuring the Fiscal Health of Local Governments: Focus on Chicago Suburban Municipalities" Urban Affairs Review, 40(1): 78-114.
Hendrick, R. M. (2008, December). The Story of the City of East St. Louis. Retrieved December 1, 2011, from http://www.uic.edu/cuppa/pa/faculty/vitae pdf/The Story of E St Louis.pdf (Archived by WebCite[R] at http://www.webcitation.org/63bu9UDLb)
Hendrick, R. M. (2011). Managing the Fiscal Metropolis: The Financial Polices, Practices and Health of Municipalities, (Washington, DC: Georgetown University Press).
Jones Day (2012). Stockton, California, Ruling: Bankruptcy Court Powerless to Prevent Retiree Benefit Reductions by Municipal Debtor. Retrieved May 24, 2013, from http://www.ionesday.com/Stockton-California-Ruling Bankruptcy-Court-Powerless-to-Prevent-Retiree-Benefit-Reductions-byMunicipal-Debtor-12-01-2012/?RSS=true
Keating, A. D. (2005). Metropolitan Growth. In Encyclopedia of Chicago. Retrieved May 24, 2013, from http://www.encyclopedia.chicagohistory.org/pages/821.html
Kincaid, J. (1989). "Fiscal Capacity and Tax Effort of the American States: Trends and Issues," Public Budgeting and Finance, 9(3): 4-26.
Kloha, P., Weissert, C. S., & Kleine, R. (2005, May/June). Developing and Testing a Composite Model to Predict Local Fiscal Distress. Public Administration Review, 65(3), 313-323.
Krane, D., Rigos, P. N., Hill, Jr., M. B. (eds.) (2001). Home Rule in America: A Fifty-State Handbook. Washington, DC: CQ Press.
Ladd, H. F. & Yinger, J. M. (1989). America's Ailing Cities: Fiscal Health and the Design of Urban Policy. Baltimore, MD: John Hopkins University.
Laughlin, A. M. (2005). Municipal Insolvencies: An Article on the Treatment of Municipalities Under Chapter 9 of the U.S. Bankruptcy Code. Municipal Finance Journal, 26(2): 37-59.
Lewis, C. W. (1994a). "Municipal Bankruptcy and the States: Authorization to File Under Chapter 9," Urban Affairs Quarterly, 30: 3-26
Lewis, C. W. (1994b). Budgetary Balance: The Norm, Concept, and Practice in Large U. S. Cities. Public Administration Review, 54(6), 515-524.
Maher, C. S. (2013). Measuring Financial Condition: An Essential Element of Management During Periods of Fiscal Stress. The Journal of Government Financial Management, 62(1): 20-25.
Matthews, R. (2013, July 18). Detroit Bankrupt: To See Detroit's Decline, Look at 40 Years Of Federal Policy. In Policy Mic. Retrieved October 27, 2013, from http://www.policymic.com/articles/45563/detroit-bankrupt-to-seedetroit-s-decline-look-at-40-years-of- federal-policv
Nowlan, J. D., Gove, S. K., & Winkel Jr., R. J. (2010). Illinois Politics: A Citizen's Guide. Urbana and Chicago, IL: University of Illinois Press.
Nollenberger, K., S. Groves & M.G. Valente (2003). Evaluating Financial Condition: A Handbook for Local Government, 4th. Washington, DC: ICMA.
Park, K. (2004). To File or Not to File Bankruptcy in the United States. Journal of Public Budgeting, Accounting and Financial Management, 16(2): 228256
Petersen, J. E. (2012, May). Stockton California's Debt Problem May Set Precedent. Governing.
Rassel, G. R. & Kravchuk, R. S. (2011). Good Debt Gone Bad: The 2007-2009 Crisis in Municipal Debt Markets. In H. Levine, J. B. Justice, & E. A. Scorsone (Eds.), Handbook of Local Government Fiscal Health (pp. 505533). Burlington, MA: Jones & Bartlett Learning.
Scorsone, E. A., & Bateson, N. (2012, February 1). Staff Paper: Evaluating a Chapter 9 Bankruptcy for City of Detroit: Reality Check or Turnaround Solution? In Department of Agricultural, Food and Resource Economics, Michigan State University. Retrieved December 19, 2012, from http://news.msu.edu/media/documents/2012/02/c763d393-5bad-4864-815b4f8ff501f289.pdf
Scorsone, E., & Wright, A. (2013, Spring). Chapter 9 as a Conduit for Restructuring Public Contracts. The Journal of Governmental Financial Management, 62(1), 50-55.
Sohl, S., Wood, C., Peddle, M, Kuhn, G., & Thurmaier, K. (2009). Measuring Financial Position of Municipalities: Numbers Don't Speak for Themselves. Public Budgeting & Finance, 29(3): 74-96.
Spiotto, J. E. (2008). The Last Resort for Financially Distressed Municipalities. In S. Feldstein and F. Fabozzi (eds.), The Handbook of Municipal Bonds (pp. 145- 180). Hoboken, NJ: Wiley and Sons.
Sweeney, A. (2013, August 11). Robbins sexual assault victim seeks answers about 1991 attack. In The Chicago Tribune. Retrieved October 10, 2013 from http://articles.chicagotribune.com/2013-08-11/news/ct-met-rape-dna-kits20130811 1 cook-county-sheriff-kits-victim
Thomson Reuters. (2012, August 14). Bankruptcy Law: Municipal bankruptcy a risk, even where it's disallowed - Fitch. In News & Insight. Retrieved December 27, 2012, from http://newsandinsight.thomsonreuters.com/Bankruptcy/News/2012/08__August/Municipal_bankruptcy_a_risk,_even_where_it_s_d isallowed_- _Fitch/ (Archived by WebCite[R] at http://www.webcitation.org/6DEuhQaf7)
United States Securities and Exchange Commission. (2011, July 29). Field Hearing on the State of the Municipal Securities Market: Distressed Communities: Remarks of James E. Spiotto. Retrieved May 13, 2013, from http://www.sec.gov/spotlight/municipalsecurities/statements072911/spiotto.pdf
Wang, X., Dennis, L., & Tu, Y. S. (2007, Summer). Measuring financial condition: A study of US States. Public Budgeting and Finance, 27(2): 1-21.
The World Bank. (2004). Intergovernmental Finance in Hungary: A Decade of Experience 1990-2000. Retrieved May 21, 2013, from http://wwwwds.worldbank.org/external/default/WDSContentServer/IW3P/IB/2005/12/07/ 000012009 20051207142137/Rendered/PDF/345770HU0rev0Intergovernmen tal0finance.pdf
Rebecca Hendrick (email@example.com)
Andrew Crosby (firstname.lastname@example.org)
Department of Public Administration
University of Illinois at Chicago
(1.) (http://www.census.gov/govs/cog/GovOrgTab03ss.html, downloaded on 4/29/13). The US Census Bureau recognizes the following five types of independent and oftentimes overlapping local units of government: counties, municipalities (incorporated), school districts, special districts to deliver specific services, and townships (only 20 states have townships).
(2.) State governments in the US are not authorized by the federal government to file for bankruptcy.
(3.) According to the US Census Bureau, the Chicago region includes counties in Illinois, Wisconsin, and Indiana. However, for purposes of this analysis it is more valid to look at municipalities in one state. Almost 80% (271) of the 347 municipalities in the region are in Illinois as of 2007, but only 265 municipalities existed in 2000.
(4.) According to Rassel and Kravchuck (2011) there were 183 defaults on debt by local governments in 2009 alone which is the "largest number since 1992, up from 162 in 2008." (p. 512).
(5.) Other criteria include that the government has a desire to implement a plan to resolve its debts and it shows willingness to negotiate in good faith with its creditors regarding an adjustment plan.
(6.) The Chicago region has more local (and overlapping) governments than any other metropolitan region and even surpasses the New York metropolitan region with only 1,535 local governments.
(7.) Counties or townships may have limited authority over land use and economic development within unincorporated areas.
(8.) Spiotto (2009) does not list Illinois as a state that allows municipal bankruptcy. However, in 2011 remarks to the U.S. Securities and Exchange Commission, Spiotto lists Illinois as a state with "limited authorization" (United States Securities and Exchange Commission, 2011, p. 62).
(9.) Home rule governments in Illinois may do anything except that which is prohibited by state law, but non-home rule governments may only do that which is allowed by state law. This distinction is similar to the classic view of home rule and non-home rule local governments in the US generally (Krane et al, 2001).
(10.) Non-home rule municipalities may not increase property taxes more than the lesser of 5% or the rate of inflation.
(11.) Three-year averages are used to reduce measurement error and the incompatibilities associated with variation in accounting practices among the suburbs. Discussions with financial managers and auditors in the region indicate that major timing differences in accounting for revenues and expenditures disappear over a three year time period. Averaging eliminates the effects of these timing differences, and reduces the impact of unique events that may dramatically affect revenues or expenditures in the short-run. For instance, financial data for 1998 represents the mean of years 1997 to 1999.
(12.) Similar to the Representative Revenue System (Kincaid, 1989), Ladd and Yinger (1989) compare individual governments to the group using the mean of the variable, although one could also use a median.
(13.) Unless they are one of the four satellite cities in the region, municipalities furthest from the City of Chicago have fewer service responsibilities (e.g. water and sewer), and municipalities in a fire district are not likely to provide separate fire services, which are costly. All per capita values were also multiplied by percent residential EAV to reduce the distortion in these measures as the jurisdictions become less residential. Many jurisdictions have a lot of commercial and industrial activity, some of which have very few residents, which makes their per capita measures exceedingly high and less accurate as a standardized value.
(14.) The purpose of this change, which does not change the relative values of the index, is to insure that all component variables in the index move in the same direction and that the index will be positive. Keeping negatively correlated component variables in the original direction creates some negative index values, which cannot be used accurately in the Revenue Wealth / Spending Needs ratio to reflect higher or lower levels of wealth relative to need.
(15.) Reliability tests, such as Cronbach's alpha, were not applied in this case because revenue wealth and spending needs are not latent variables. Analysis also shows that using Z scores rather than percent medians in the calculation of these indices increases their predictive validity, but Z scores cannot be used accurately in ratios, such as wealth divided by need, due to their negative values.
(16.) We also examined correlations between all five solvency indicators and the government bond ratings by Standard & Poor's for 2009. Of the 105 municipalities with bond ratings in the region, the wealth / need index correlated with the bond rating at r =.58. Correlations with bond ratings and the other four solvency indicators were greater than .3.
(17.) From a full accrual perspective, the value of assets minus liabilities is called equity or net income. See Appendix A for a discussion of the accounting framework used by most governments in this region.
(18.) The number of governments that are classified into the cells in Table 1 is higher than what is shown in the designated cells in Table 5 because we include governments in Table 1 that are in the same locations in crosstabulations for fiscal bad times and the same locations in crosstabulations (good times and bad times) of wealth relative to need and the other short-term solvency index.
(19.) The Financially Distressed Cities Law (65 ILCS 5/Art. 8 Div. 12) that applies to home rule municipalities, and the Local Government Financial Planning and Supervision Act (50 ILCS 320/) that applies to other non-home rule governments, except school districts, do not allow the state to unilaterally intervene in the fiscal affairs of the local governments. Rather, the state must be invited in by the local government via the passage of a local ordinance or a supermajority vote of the governing body respectively.
Table 1. Relation between Short-Term and Long-Term Solvency SHORT LONG TERM SOLVENCY TERM SOLVENCY LOW HIGH LOW A: Fiscal crises C: Poor financial expected management and bad financial decisions HIGH B: Good financial D: Fiscal crises management and sound not expected financial decisions Table 2. Percent Change in Three Solvency Indicators: 1998-2009 1998 1999 2000 2001 2002 2003 - - - - - - 1997 1998 1999 2000 2001 2002 Percent Change Fund Balance as Percentage of Expenditures (1) mean 11.2 -1.5 5.7 -20.6 -42.5 75.6 median 3.3 1.4 1.8 -1.8 -9.5 -8.0 Percent Change Days of Cash on Hand (1) mean 12.9 7.6 5.7 5.5 -0.6 -0.6 median 6.8 5.3 2.7 2.1 -3.6 -4.5 Percent Change Total Revenues (1,2) mean 5.1 6.2 5.7 2.2 0.5 0.5 median 2.7 3.8 3.8 0.9 -1.0 -1.3 2004 2005 2006 2007 2008 2009 - - - - - - 2003 2004 2005 2006 2007 2008 Percent Change Fund Balance as Percentage of Expenditures (1) mean -24.8 25.5 -9.8 29.2 -21.1 -62.7 median -2.6 7.1 2.4 1.4 -6.5 -4.7 Percent Change Days of Cash on Hand (1) mean 20.2 4.6 6.1 1.7 -2.9 -2.5 median -1.9 3.7 2.7 -0.4 -3.6 -3.3 Percent Change Total Revenues (1,2) mean 2.5 11.9 4.9 3.3 -1.1 -3.5 median 2.1 2.8 3.2 2.0 -0.7 -2.9 (1:) capital funds excluded (2:) corrected for inflation Table 3. Correlations between Fiscal Good Times and Bad Times for the Solvency Indices Moderate All Years Years Removed Days of Cash on r = 0.82 r = 0.81 Hand p = .000 p = .000 Expenditure r = 0.76 r = 0.73 Coverage p = .000 p = .000 Revenue Burden r = 0.96 r = 0.92 p = .000 p = .000 GO Debt Burden r = 0.86 r = 0.72 p = .000 p = .000 Table 4. Correlations between Solvency Indices for Fiscal Good Times Wealth / GO Debt Revenue Spending Need Burden Burden Coverage Days of Cash Pearson r = 0.14 -0.25 -0.27 0.96 on Hand Prob. = 0.026 0.000 0.000 0.000 Spending Pearson r = 0.15 -0.26 -0.28 Coverage Prob. = 0.017 0.000 0.000 Revenue Pearson r = -0.45 0.434 Burden Prob. = 0.000 0.000 GO Debt Pearson r = -0.19 Burden Prob. = .001 Table 5. Crosstabulation Table of Septiles of Long-Term and Short-Term Solvency Indices LONG-TERM SOLVENCY: Revenue Wealth / Spending Needs 1 2 3 4 SHORT-TERM 1 15 5 7 4 SOLVENCY: 39.5% 13.5% 17.9% 10.8% Spending Coverage 2 7 5 4 5 18.4% 13.5% 10.3% 13.5% Good Times 3 5 6 6 4 13.2% 16.2% 15.4% 10.8% 4 2 9 3 8 5.3% 24.3% 7.7% 21.6% 5 4 3 6 8 10.5% 8.1% 15.4% 21.6% 6 2 3 8 5 5.3% 8.1% 20.5% 13.5% 7 3 6 5 3 7.9% 16.2% 12.8% 8.1% Total 38 37 39 37 100.0% 100.0% 100.0% 100.0% LONG-TERM SOLVENCY: Revenue Wealth / Spending Needs 5 6 7 Total SHORT-TERM 1 2 2 3 38 SOLVENCY: 5.3% 5.3% 7.9% 14.3% Spending Coverage 2 6 7 4 38 15.8% 18.4% 10.5% 14.3% Good Times 3 6 5 5 37 15.8% 13.2% 13.2% 14.0% 4 5 7 5 39 13.2% 18.4% 13.2% 14.7% 5 7 4 5 37 18.4% 10.5% 13.2% 14.0% 6 7 5 8 38 18.4% 13.2% 21.1% 14.3% 7 5 8 8 38 13.2% 21.1% 21.1% 14.3% Total 38 38 38 265 100.0% 100.0% 100.0% 100.0%
|Printer friendly Cite/link Email Feedback|
|Author:||Hendrick, Rebecca; Crosby, Rebecca|
|Publication:||Public Finance and Management|
|Date:||Jan 1, 2014|
|Previous Article:||Default without capital account: the economics of municipal bankruptcy.|
|Next Article:||Bankruptcy: the divergent cases of the city and the county of San Bernardino.|