After the Storm: Investigating Ohio’s Recovery from the ...



The Ohio State University John Glenn College of Public AffairsAfter the Storm: Investigating Ohio’s Recovery from the Great Recession and its Impact on Local Capital InvestmentKate Lewis-LakinA policy paper submitted in partial fulfillment of the Masters in Public Administration degreeSpring 2016Executive SummaryThe Great Recession of 2008 and 2009 hit both state and local governments in Ohio hard in terms of decreased revenue from income, sales, and property taxes. Facing its own budget crisis, the state government significantly cut the state’s primary revenue sharing program, the Local Government Fund. These cuts amounted to a 50 percent reduction from 2001 to 2013, adding to existing strain on local government budgets. This project seeks to further understand impacts of the Local Government Fund cuts on local government through analysis of quantitative budget data from 2007-2013. Revenue-sharing programs from state to local government fall under the study of fiscal federalism, the vertical structure of government financing. Over the past decade, researchers of fiscal federalism have seen movement towards decentralization, with revenue-sharing cut in order to increase the reliance of local governments on funding collected from within their own jurisdictions. When local governments are faced with cuts to revenue sharing, they are forced to make hard decisions about raising local taxes versus cutting spending, known as cutback management. In Ohio, local governments have grappled with these cuts to revenue sharing in multiple ways. A survey conducted in 2013 began to shed some light on the responses of local government leaders to the policy changes. On the revenue side, local governments have raised charges for service and other fees in addition to raising property and income taxes where possible. The most frequent expenditure changes were cuts to capital spending. Conversations with local government leaders across the state solidified and clarified the impact of these cuts on capital investment. The two primary research questions under investigation are 1) How did local government revenues change as a result of cuts to the Local Government Fund? and 2) How did these changes to revenue impact local governments’ ability to invest in capital projects? Analysis has been conducted on budget data from 250 cities measuring total revenues and expenditures, key revenue sources, and capital outlays from 2007-2013. These data were collected from the Ohio Auditor of State and Department of Taxation. Key demographic variables from the United State Census Bureau are also included to capture differences between municipalities. Descriptive analysis was used to answer the first research question. Overall, cities have seen decreasing property tax revenues and increasing income tax revenues since the recovery began in 2009. These increased income tax revenues have been due in part to increasing municipal income tax rates, with a statewide average rate increase of 0.48 percentage points. Capital outlays have declined since the Great Recession and through changes to the Local Government Fund, with a 22 percent decrease in the statewide average from 2007 to 2013. Regression analysis was used to examine whether certain years or changes to revenue sources had a significant impact on capital outlays. It was expected that capital outlays would have a positive relationship to Local Government Fund revenues and a negative relationship in the years after the cuts took place. The model found that capital outlays were indeed significantly lower in the years after the Local Government Fund was reduced. These results offer tentative confirmation that cuts to the Local Government Fund have led to declining capital investment in cities across Ohio. Ohio’s local government leaders may use this information as well as their own experiences to begin lobbying state legislators for a dedicated capital improvement fund for local governments. Due to limitations to the data and to the model, further research should be pursued to confirm and expand upon these findings.AcknowledgementsThis project would not have been possible without the support of many in the Glenn College and in Ohio’s state and local governments. Mayors Debbie Sutherland, Mike Summers, and Nan Whaley are to thank for beginning the conversations that served at the catalyst for this project, and for continuing to provide guidance and insight throughout the research process. A dinner conversation with John Begala, Gene Krebs, and Larry Long contributed invaluable information for further developing the study. Sharon Hanrahan at the Ohio Auditor of State’s office was integral in locating the data to make this study possible. I thank all of these individuals for their support of my work and continued research on such an important topic.Ted Staton provided valuable insight and feedback throughout my research process and has served as my local government mentor since May 2015. Ted’s dedication to the development of future local government leaders is unmatched. I have been lucky to work with such a distinguished city manager that cared so much about my development and success as a professional. Thank you for supporting me and inspiring me to pursue such a rewarding line of work.Thank you to John Delia for your valuable assistance and support for this research, and I wish you the best in continuing with this project. Finally, I wouldn’t have gotten further than a question without the constant help of my research mentors, Dr. Ned Hill, Dr. Charlotte Kirschner, and Dr. Rob Greenbaum. Thank you for answering all my emails, talking me through regression analysis, and supporting me in every step along the way. Table of ContentsIntroduction ……………………………………………………………………………………… 5Policy Background ………………………………………………………………………………. 8Figure 1: Basis of Funding for the Local Government Fund ……………………………. 9Local Policy Alternatives ………………………………………………………………………. 12Key Stakeholders ………………………………………………………………………………. 14Literature Review ……………………………………………………………………………..... 16Methods ………………………………………………………………………………………… 24 Data …………………………………………………………………………………………….. 27Figure 2: Descriptive Statistics ………………………………………………………… 28Results ………………………………………………………………………………………….. 28Figure 3: Statewide Average – Total Expenditures and Revenue per Capita ………….. 29Figure 4: Statewide Average – Income Tax per Capita ………………………………... 30Figure 5: Statewide Average – Municipal Income Tax Rate …………………………... 31Figure 6: Statewide Average – Revenue Sources per Capita ………………………….. 32Figure 7: Statewide Average – Capital Outlays per Capita ……………………………. 33Figure 8: Statewide Average – Percent of Total Revenues ……………………………. 35Figure 9: Regression Analysis …………………………………………………………. 36Figure 10: Correlation Matrix ………………………………………………………….. 41Figure 11: Variance Inflation Factors ………………………………………………….. 42Conclusions …………………………………………………………………………....……….. 43Bibliography …………………………………………………………………………………… 47Appendix A: Breakdown by Population ……………………………………………………….. 49Figure 12: Local Government Fund Distributions as Percent of Total Revenue, by Population ………………………………………………………………….. 50Figure 13: Property Taxes as Percent of Total Revenue, by Population ………………. 51 Figure 14: Income Taxes as Percent of Total Revenue, by Population ………………... 52Figure 15: Charges for Services as Percent of Total Revenue, by Population ………… 53Figure 16: Fees, Licenses and Fines as Percent of Total Revenue, by Population …….. 53Appendix B: Breakdown by Household Income ………………………………………………. 54Figure 17: Local Government Fund Distributions as Percent of Total Revenue, by Median Household Income ………………………………………………… 55Figure 18: Property Taxes as Percent of Total Revenue, by Median Household Income ……………………………………………………………………… 55 Figure 19: Income Taxes as Percent of Total Revenue, by Median Household Income . 56Figure 20: Charges for Services as Percent of Total Revenue, by Median Household Income ……………………………………………………………………… 57Figure 21: Fees, Licenses and Fines as Percent of Total Revenue, by Median Household Income ……………………………………………………………………… 57IntroductionThe Great Recession of 2008 had a dramatic impact on funding at all levels of government. With severe unemployment and falling wages, federal, state and local governments experienced substantially decreased income tax revenues. Consumer purchasing fell, contributing to nationwide drops in sales tax revenues. With some delay due to assessment cycles, property tax revenues also fell, the impact of which was especially hard on local governments. As all levels of governments struggled with these diminished tax revenues, intergovernmental distributions were also cut. In Ohio, this was most clearly exhibited by a 50 percent decrease in funding to the Local Government Fund, the state’s revenue sharing program for local governments. Total distributions to cities and villages from this program equaled $374 million before the cuts were instituted; this fell to $262 million in 2012 and $201 million in 2013 (Department of Taxation, 2016). In justifying these changes to the Local Government Fund, state leaders point to the comparatively high amount of support that Ohio’s state government was giving to local governments prior to the Great Recession. Local governments in Ohio also have the ability to raise their own revenue through a municipal income tax, an option that is not available in most states and is more widely utilized in Ohio than elsewhere. Local government leaders have expressed concern about the impact of these cuts on their ability to provide necessary services to their communities. These effects were compounded by the elimination of the estate tax and changes to the tangible personal property tax. Mayors and managers were faced with hard decisions to cut spending, increase locally collected revenues, or both. The goal of this research is to define and measure the impacts of reducing the Local Government Fund on local governments. The first research question under investigation in this paper is how Ohio local governments adjusted revenue sources between 2007 and 2013 as a result of the Great Recession and subsequent state policy changes. A second research question is how these changes in revenue sources have impacted capital investments at the local level. These questions come from the results of a 2013 survey of municipal governments in Ohio, as well as conversations with mayors and city managers across the state. These two sources indicate that reducing capital investment was the most commonly utilized change in expenditures, and that this reduction in capital investment may have a long-term impact on the strength of infrastructure across the state. Based on the findings of the 2013 survey and trends of cutback management across the country, it is expected that local governments in Ohio will show increased revenues from charges and fees, income taxes, and property taxes in response to decreased revenues from state revenue sharing programs. It is also expected that capital expenditures have declined, and that this decline has been especially pronounced after the cuts to the Local Government Fund were instituted. Statistical analysis is used to test these hypotheses. First, descriptive analysis is conducted in order to more fully understand the ways in which revenue streams across the state changed from 2007 to 2013. This time period captures both changes instituted after the Great Recession and those that emerged after the reduction of the Local Government Fund. Second, regression analysis is used to study and measure the extent to which these changes to revenue streams have impacted capital investment by local governments. The analysis is based primarily on budget data reported to the Auditor of State by 250 Ohio cities from years 2007 through 2013. Data from the Ohio Department of Taxation of Local Government Fund distributions per city and estate tax collection are also included. Counties, townships, and villages are excluded from the analysis at this time. Additional data from the United States Census Bureau are used to analyze whether the impacts of these policy changes may have varied based on population or median household income of a community. The results of the analysis indicate that capital outlays did significantly decrease from 2010-2013, after the Great Recession and changes to the Local Government Fund. In 2010, capital outlays per capita were 21.2 percent lower than in 2007. This difference increased in the following years, with capital outlays per capita 30.1 percent lower in 2011, 33.3 percent lower in 2012, and 26.2 percent lower in 2013. Further research is needed in order to better capture the specific impact that the Local Government Fund may have had on capital investment. Going forward, the findings of this and future research will be beneficial to local government leaders as they make spending decisions within their own jurisdictions. It may also be beneficial in the development of a new state revenue-sharing program specifically for local capital investment. It is recommended that Ohio’s local government leaders pursue further research on this topic in order to advance their pursuit of this capital fund. The first section of this paper gives the history of Ohio’s Local Government Fund and details the changes to local government financing that occurred after the Great Recession. The two sections following consider local policy alternatives and key stakeholders for these findings. Next, a literature review offers background to the study of fiscal federalism and cutback management, as well as how these concepts apply to the decisions made in Ohio at the state and local level. The data and methods section offers further information on the data collected and the model used for analysis. The paper concludes with results of descriptive and statistical analysis, conclusions, and directions for further research. Policy BackgroundHistory of the Local Government FundThe Ohio General Assembly established the Local Government Fund (LGF) in 1934 in conjunction with the first state sales tax. At the time, local governments were experiencing fiscal strain as a result of decreased tax revenues during the Great Depression. In its first year, about 40 percent of state sales tax revenues went to local governments. Until 2011, the Local Government Fund was financed by a set percentage of revenues from a variety of state funding streams. Since its inception, the funds have been distributed to municipalities in two ways. One portion is distributed in undivided funds to counties and county budget authorities determine the distribution of funds to the individual cities, villages, and townships. A portion of the funds is also distributed directly to municipalities that levy an income tax (Taxation, 2005). Policy ChangeIn 2008, declining tax revenues due to the Great Recession caused budget distress at the state level. As these effects were fully realized, reductions to the Local Government Fund were instituted in order to balance the state budget. There is an irony here since the Local Government Fund was originally established to help local governments during a period of economic depression and was later cut in a similar economic environment. Prior to these changes, the amount of money in the Local Government Fund was determined by a percentage of the state’s General Revenue Fund from the preceding month. In fiscal year 2011 (July 2010 – June 2011), the Local Government Fund was equal to 3.68 percent of the state General Revenue Fund. This was changed in the fiscal year 2013 biennial state budget (July 2011 - June 2013), and these changes were continued in the fiscal year 2015 state budget (July 2013 - June 2015). The changes are described below and further detailed in Figure 1 on page 9 (Taxation, 2013). The new formula temporarily changed the basis of funding for the Local Government Fund. For fiscal year 2012 (July 2011-June 2012), monthly Local Government Fund financing was equal to 75 percent as the same month in fiscal year 2011. In fiscal year 2013 (July 2012 – June 2013), this was reduced to 50 percent of same month Local Government Fund in fiscal year 2011. The fiscal year 2015 budget returned to a funding formula based on a percentage of the state’s general revenue fund. The funding percentage was determined by dividing fiscal year 2013 Local Government Fund deposits by the total state General Revenue Fund. This essentially continued the 50 percent reduction from the beginning of these changes, working out to 1.66 percent of the General Revenue Fund, as opposed to 3.68 percent before the cuts were instituted. These details are also shown in Figure 1 below (Taxation, 2013). Figure 1: Basis of Funding Changes to Local Government FundMonthCalendar Year 2011CY 2012CY 2013CY 2014January3.68% of Dec. 2010 General Revenue Fund (GRF)75% of Jan. 2011 LGF50% of Jan. 2011 LGF1.66% of Dec. 2013 GRFFebruary3.68% of Jan. 2011 GRF75% of Feb. 2011 LGF50% of Feb. 2011 LGF1.66% of Jan. 2014 GRFMarch3.68% of Feb. 2011 GRF75% of Mar. 2011 LGF50% of Mar. 2011 LGF1.66% of Feb. 2014 GRFApril3.68% of Mar. 2011 GRF75% of Apr. 2011 LGF50% of Apr. 2011 LGF1.66% of Mar. 2014 GRFMay3.68% of Apr. 2011 GRF75% of May 2011 LGF50% of May. 2011 LGF1.66% of Apr. 2014 GRFJune3.68% of May 2011 GRF75% of June 2011 LGF50% of June 2011 LGF1.66% of May 2014 GRFJuly3.68% of June 2011 GRF50% of July 2010 LGF50% of July 2010 LGF1.66% of June 2014 GRFAugust75% of Aug. 2010 Local Government Fund (LGF)50% of Aug. 2010 LGF1.66% of July 2013 GRF1.66% of July 2014 GRFSeptember75% of Sep. 2010 LGF50% of Sep. 2010 LGF1.66% of Aug. 2013 GRF1.66% of Aug. 2014 GRFOctober75% of Oct. 2010 LGF50% of Oct. 2010 LGF1.66% of Sep. 2013 GRF1.66% of Sep. 2014 GRFNovember75% of Nov. 2010 LGF50% of Nov. 2010 LGF1.66% of Oct. 2013 GRF1.66% of Oct. 2014 GRFDecember75% of Dec. 2010 LGF50% of Dec. 2010 LGF1.66% of Nov. 2013 GRF1.66% of Nov. 2014 GRFSource: Department of Taxation, 2013 (Shading reflects state budget cycles)Why cut the Local Government Fund?A number of factors motivated state policymakers in Ohio to cut the Local Government Fund. In 2010, the Center for Community Solutions put forth a comprehensive set of solutions for Ohio’s fiscal crisis, including reducing and restructuring the Local Government Fund. Ohio, the report argued, is one of only a few states in the nation with a widely used municipal income tax, giving local governments a greater ability to raise tax revenues directly from their residents. Of the 940 cities and villages in Ohio that receive distributions from the Local Government Fund, 540 also assess a municipal income tax (Begala, 2010). Compared to other states, Ohio has a pattern of relatively high local tax collection and low state tax collection (Begala, 2010). The Center for Community Solutions argued that reductions to the Local Government Fund would have a major impact on reducing the state deficit. A 10 to 20 percent reduction in combined municipal and county shares of the Local Government Fund was projected to produce a savings of about $132 to $264 million over the 2012-2013 biennium (Begala, 2010). To put this in perspective, the state budget deficit in 2011 was speculated to be anywhere between $5.9 to 8 billion (Marshall, 2011). The writers also argued for the replacement of the Local Government Fund with new, targeted Local Government Collaboration Grants that would encourage consolidation and collaboration between local governments in order to improve efficiency (Begala, 2010). Changes to the Estate Tax and Tangible Personal Property TaxCutting the Local Government Fund was not the only change to municipal government financing in Ohio after the Great Recession. Ohio’s cities have also felt the effects of the elimination of the estate tax and phase-out of distributions from the tangible personal property tax. These changes went into effect around the same time as the cuts to the Local Government Fund, compounding the fiscal stress experienced by cities across the state. Though the Local Government Fund is the primary policy change under investigation in this study, understanding these two additional polices contributes to a fuller picture of the fiscal environment.The Ohio General Assembly enacted the estate tax in 1968, replacing a state inheritance tax that had been in effect since 1893. In state fiscal year 2010, the tax applied only to those properties with a net taxable value greater than $338.3 thousand. Net taxable value above this threshold but below $500 thousand was taxed at a 6 percent rate, and any value over $500 thousand was taxed at a 7 percent rate. Most of the revenues from the estate tax were distributed to the local governments where the estates were located, with a small remainder put into the state General Revenue Fund. Revenues from the estate tax equaled $285.8 million in state fiscal year 2010, and about $230.8 million of this was distributed to local governments (Testa, 2010).In 2011, the Ohio General Assembly repealed the estate tax for estate owners with a date of death on or after January 1, 2013. With this policy change, a revenue stream was completely eliminated from the budgets of local governments. While some municipalities still received estate tax revenues from outstanding claims during the 2013 calendar year, these revenues fell significantly by 2014 and have since diminished. Because the scope of this study is from 2007 through 2013, effects of the estate tax are not yet apparent in the data (Testa, 2010).The tangible personal property tax is a tax assessed on items of major value that are not physically affixed to the land. Beginning with House Bill 66 in 2005, a five-year phase-out of the tangible personal property tax began. This phase-out included a system of replacement payments to local governments and school districts from the state in an attempt to offset some revenues lost in this policy change. The phase-out was complete for most taxpayers by 2008. Significant changes were made to the replacement payment system in 2011, including cuts to both local government and school district distributions (Taxation, 2016). Local Policy AlternativesFaced with declining revenues on all fronts, local governments in Ohio have been faced with the tough decision of increasing revenues, cutting expenditures, or some combination of both. This section will review these options as they relate to the central research questions of the paper. Each method has its own benefits and drawbacks and is governed by certain state regulations. Policy alternatives to consider at the state level will be discussed in the conclusion of the paper. In Ohio, municipal income tax rates are determined by a vote of those living within the local jurisdiction. The maximum income tax rate that can be imposed without a vote is one percent (Taxation, 2011). Any higher rate or change to the income tax rate is required to be put on the ballot and approved by a majority of voters. Because income taxes are collected from non-residents who work within a municipality, they are usually more beneficial to large cities with a high number of employers. Cities must also consider the needs of businesses when changing income tax rates. Having a higher income tax rate than neighboring communities could be detrimental to a city that hopes to attract new businesses. Often in metropolitan areas, the tax rate of the central city is treated as the maximum possible for suburban communities. Columbus, for example, has a 2.5 percent income tax rate, and none of its suburbs have attempted to exceed that rate. Changing property tax rates is even more complicated. Ohio law allows only a 10 mills property tax, equal to one percent, to be assessed without a vote. This rate includes property tax revenues that will go to the public school district, county, and public library as well as the city government. Voters must approve higher property tax rates or changes above 10 mills. Like the income tax, this dependency on election outcomes makes changing property tax rates a difficult endeavor. The ability of a city to raise property tax rates is also complicated by the number of governments within a jurisdiction that share property tax revenues and that also pursue increases to the property tax. The ability of a city to raise revenue from charges and fees is shaped by a number of factors. Revenue is based both on demand and utilization of the services as well as by the fees that are set for those services. Some charges and fees are set by city ordinance, requiring more process to change, while others are at the discretion of local administrators. Even if fee amounts are changed, changes in demand for the services could impact whether more revenue is actually collected. Charges for services also tend to have a disproportionately negative affect on low-income individuals who also tend to have the greatest need for services.Changes to expenditures in order to reduce spending are also a major policy decision made at the local level. It is not surprising that in Ohio, the choice to reduce capital outlay in response to falling revenues has been especially popular. Capital projects are long-term and often easier to push down the road than cutting current services or laying off employees. Capital projects require a significant front-end investment, so delaying them until more resources are available is often preferable to reducing the scope of the project. This is supported by a review of the literature in later sections. Key StakeholdersThe impetus for this research comes from a series of conversations with local government leaders in Ohio. Mayors and city managers know of the actions they have taken to manage fiscal stress within their own cities but are interested in better understanding statewide trends. Additionally, much of the research on these issues has been conducted by non-profit organizations with an ideological lean. Unbiased academic research will be helpful to these managers and mayors as they seek to understand larger trends and how their own cities fit in. Going forward, local government leaders will be able to use the information from this study for future lobbying efforts at the state level. If further cuts to local government revenue sharing are considered in the next state budget cycle, local government leaders will be prepared with the necessary information on the impacts that these cuts have already had on municipal governments. The second major stakeholder group for this research is state policymakers. Members of the Kasich administration and of the Ohio General Assembly were champions of the decision to cut the Local Government Fund. Research such as this that better defines the impacts of these changes is beneficial for policy evaluation and future policymaking. However, the political implications here are complicated. Governor Kasich is currently in the midst of a presidential race and regularly discusses his history in turning the state budget deficit into a surplus, and cutting the Local Government Fund was a piece of how this was achieved. Still, it is important for state policymakers to have a better understanding of these impacts, especially as local government leaders may use the findings of this research in lobbying efforts.Taxpayers are another group of stakeholders affected by these policy changes and possibly interested in the outcomes of this research. Beginning in 2013, the state personal income tax was reduced by 10 percent over three years, while the state sales tax rate was increased from 5.5 to 5.75 percent (Taxation, 2013). However, municipalities across the state increased income tax collection at the local level after cuts to the Local Government Fund were instituted in 2011. Further research will be required to examine whether taxpayers have found themselves with a net gain or loss in tax burden. Taxpayers may also be closely affected by reductions in capital spending at the local level. The first responsibility of many local governments is to maintain the physical infrastructure in their jurisdiction; when potholes go unfilled, local governments are the first to blame. A final group of stakeholders to consider is the business community in Ohio. Businesses have faced the same changes to state taxes as individual residents, plus an additional tax reduction on business income for small businesses. However, it remains to be seen whether tax increases at the local level may be affecting the net change in tax burden experienced by Ohio’s businesses. Municipal income taxes have a particular effect on businesses because they can affect where a business chooses to locate. If a small business experiences an increase in municipal income tax rate in their present location, they may consider moving to a different jurisdiction with a lower tax rate. Local capital investment also affects businesses – if the capital infrastructure in a certain jurisdiction does not suit the needs of a certain business, this may also be a reason to relocate to a different city or state.Literature ReviewThe following literature review investigates two questions that are central to this study. The first portion examines why the Local Government Fund was reduced. Theories of fiscal federalism and the benefits principle argue that providing and funding services at lower levels of government is more efficient, but it also create concerns about equity. The second questioned uses literature of cutback management to examine the decisions that local governments make when faced with declining revenues. Theory and survey research in Ohio reveal that increasing user fees is common on the revenue side, whereas reducing capital investment is frequently used to manage expenditures. Fiscal FederalismThe study of distributions from state to local governments falls under the broader umbrella of “fiscal federalism.” Fiscal federalism is the study of the vertical fiscal structure of the public sector (Oates, 1999). Federal systems constantly make decisions about the proper level of government to provide certain services, as well as which level of government should fund those services. The tools of fiscal federalism include direct taxation from all levels of government as well as intergovernmental grants between federal, state, and local levels. The primary decision point for a federal system is the alignment of responsibilities and revenue sources with the proper levels of government (Oates, 1999). The principles of fiscal federalism can make a compelling point for providing and funding services at the local level. If the consumption of a service is limited to a particular jurisdiction, it is more efficient to provide that service at the local level, because the service can be tailored more directly to the needs of that specific population. Following the benefits principle of taxation, services provided at the local level should also be funded by taxes collected at the local level. This ensures that only those individuals who are utilizing a service are paying for that service, and further increases efficiency because tax dollars are spent at the same level of government that they are collected (Oates, 1999). There are two major drawbacks to this model. First, local governments do not actually have the economic control required to ensure that services perfectly fit the needs of their jurisdiction. This is due to the comparative ease of residents to move between jurisdictions. This ease of movement decreases at higher levels of government; it is easier to move between cities than states and between states than countries. Local governments are, to an extent, beholden to the practices of their neighboring municipalities because of this potential mobility of residents and businesses (Oates, 1999).There is also an equity issue in providing and funding government services at a local level. Local governments with larger and wealthier tax bases will be able to provide better services, attracting more residents and businesses that will in turn pay taxes for even better services. The opposite is true in poorer municipalities. Lower tax collection will lead to poorer services, which will drive residents and businesses away from these areas and into those with better services. This further reduces the local government’s ability to provide adequate services, further depressing these areas and increasing the disparity between municipalities (Oates, 1999).Intergovernmental grants like the Local Government Fund serve an important role in a fiscal federal system. It is realistically impossible to limit the benefits from a public service to a single jurisdiction. Intergovernmental grants serve to internalize some of these spillover benefits by providing funding from higher levels of government for services having cross-jurisdictional benefits. These grants also reduce the equity issues discussed above by redistributing resources among poorer and richer jurisdictions across the state. Finally, intergovernmental grants can also lead to an improved overall tax system. Local tax collection becomes quickly complicated for individuals who work in a separate jurisdiction from where they work and for businesses that operate across multiple jurisdictions. Collecting taxes at the state level and redistributing the revenues to local governments prevents these individuals and businesses from having to pay taxes separately in each jurisdiction (Oates, 1999).The “New Normal”Across the nation from 1957 to 1986, centralization of finances at the state level increased, with state revenue collection making up a greater portion of combined state and local revenues (Pagano and Mullins, 2005). Intergovernmental revenue from states to all local governments doubled in constant dollars from 1980 to 2002 (Pagano and Mullins, 2005). More recently, local governments across the country have seen declining distributions from their state governments in the face of economic hardship. In the early 2000s, states began to cut local government aid, a trend that Hoene and Pagano (2003) named “fend-for-yourself federalism.”This aligns with trends in the literature towards decentralization in fiscal federalism in order to achieve gains in efficiency described above. The Great Recession that began in 2008 has increased reductions in state aid to local governments. State governments expect negative effects from this recession for at least ten years (Scorsone and Plerhoples, 2010). As states face budgetary strain, intergovernmental grants to local governments tend to be even further negatively affected. The effect of the recession directly on local governments is still being felt due to delays in property tax assessment cycles. Even now, many years after 2008, state and local governments continue to operate under restrained fiscal conditions, inspiring speculation of a “new normal” based on this model of “fend-for-yourself federalism” (Scorsone and Plerhoples, 2010; Hoene and Pagano, 2003). Martin, Levey and Cawley (2010) undertook a national study to investigate whether local governments can expect to return to a pre-recession state of operation or whether the recession has fundamentally changed fiscal operations. Their findings suggest that the latter may be true. The National League of Cities found that municipalities across the country ended 2010 with the largest year-over-year general fund reductions in the last 26 years (Martin, Levey and Cawley, 2012). In a 2011 survey of counties, decreased state funding was one of the most commonly cited reasons for budgetary shortfalls at the local level (Martin, Levey and Cawley, 2012). Long-term trends may also be at play in the creation of this new normal. Prior to the 2008 recession, local governments nationwide may have been on an unsustainable pattern of growth. In most spending categories, local government expenditures had increased at a rate that outpaced both population growth and inflation (Martin, Levey and Cawley, 2012). The fiscal reductions that resulted from the 2008 recession may have brought local governments to a new, more sustainable level. This argument was used in Ohio to advocate for reductions to the Local Government Fund. Either way, local governments face a “new normal” characterized by restrained revenues and expenditures that is expected to remain for years to come. Cutback ManagementWhen faced with declining revenues on all fronts, including state support, how do local governments respond? Answering this question requires a broader investigation into the scholarship of cutback management as it applies to local governments. Cutback management emerged as an area of study in the 1970s, when for the first time since World War II cities and states were forced to address revenue constraints on a major scale (Scorsone and Plerhoples, 2010). Governments face a basic decision in cutback management: to increase revenues, cut expenditures, or some combination of both. From the 1970s through the early 2000s, government tended to respond to fiscal crises by both increasing taxes and cutting services. Since the economic downturn in 2001, the local government response has shifted primarily to service cuts, with a decreasing reliance on tax increases. Scorsone and Plerhoples (2010) offer a number of reasons why this may be true. State and local governments have tended in the past decade to not cut taxes during periods of economic prosperity as was typical prior to 2000. Because of this, many governments have been able to build up and rely on their reserves during the current fiscal crisis rather than instituting tax increases (Scorsone and Plerhoples, 2010). Another common practice has been to increase charges for services and other fees. According to results from an annual survey by the National League of Cities, from 2008 to 2010 at least one in five cities reported increases in the number of fees levied (Scorsone and Plerhopes, 2010). Morgan and Pammer (1988) examined strategies employed by cites to manage fiscal retrenchment from 1980-1983. They found that 87.3 percent of cities surveyed increased user fees, while 76.5 percent sought new local revenues sources (Morgan and Pammer, 1988). Local governments also commonly respond to economic hardship by cutting capital investment. This is well supported in the literature. Cities tend to defer capital investment in favor of meeting current service delivery obligations and keeping personnel (Pagano, 2002). Morgan and Pammer (1988) found that 60.3 percent of cities studied reduced capital expenditures. Pagano (2002) found that growth in three key revenue sources – own-source revenues, debt, and intergovernmental revenue – during the economic boom from 1993-2000 allowed cities to increase their capital expenditures. All three of these revenue streams fell substantially for Ohio’s cities during and after the Great Recession, so it follows that capital expenditures would also be expected to fall (Pagano, 2002). Jimenez (2009) applied the findings of these two studies to state governments, and found that in times of fiscal stress state governments will tend to reduce their share of spending for local capital investment.The literature summarized here provides a theoretical background for fitting the changes to municipal financing in Ohio in a broader context. States across the nation have moved towards fiscal decentralization, potentially sacrificing equity for efficiency. As cities grapple with the reduction of resources on all fronts, strategies to raise revenue and decrease expenditures are both utilized. On the revenue side, increases to charges for services has become more common in recent years as opposed to increasing taxes. In regard to expenditures, deferring capital investment is an oft-used strategy because this can allow a city to maintain service levels and avoid having to cut personnel. This research seeks to take these larger trends and examine how they may apply specifically to Ohio in the wake of the Great Recession and changes to the Local Government Fund. Ohio has a diverse mix of industries and types of cities, making it likely to match national trends. However, Ohio is unique with a wide utilization of a municipal income tax, and this may lead to key differences in fiscal decisions at both the state and local level. Further study is needed on Ohio specifically to understand how responses may have been similar or different from other states across the country. The next section will examine one such study conducted in 2013, and how those findings will be used to develop the hypotheses and models used in this research. Two case studies from discussions with Ohio mayors will also be examined to emphasize the need for continued research in this area. Ohio Municipal Fiscal Assessment SurveyIn the summer of 2013, the cities of Upper Arlington, Westerville, and Loveland along with the Ohio Municipal League and the Ohio City Managers Association conducted the Municipal Fiscal Assessment Survey to examine how Ohio municipalities had responded to fiscal changes. The focus of the survey was on changes to the Local Government Fund, elimination of the estate tax, and the Great Recession. Analysis was conducted on responses completed by chief administrative officers or finance directors from 114 Ohio municipalities including cities, townships and villages. The results of this survey begin to show the impact of these fiscal changes and the way in which municipalities in Ohio responded (City of Upper Arlington, 2013). In consideration of how cities changed their revenue sources, the survey found increasing user fees to be the most common response. This was followed by increases to income tax rates and property tax rates. This echoes the findings of the literature. Another key finding was that municipalities with smaller budgets and populations saw larger general fund decreases over the time frame in consideration. This is supported by the knowledge that cities with more residents have a much larger tax base for both income and property tax collection. The survey also took an in-depth look at expenditure changes that cities made in response to the changing fiscal environment. Nearly 70 percent of municipalities in the study reduced capital expenditures and over 50 percent reduced service levels. Employee layoffs were also used by about a quarter of respondent municipalities (City of Upper Arlington, 2013).There are limitations to the methods used in this survey that open clear avenues for continued research on this topic. The sample for the survey was collected using convenience sampling, which limits the external validity of the findings. All budgetary data included was self-reported, so there may be discrepancies in how different local governments reported information. Finally, most of the data was qualitative, limiting the extent to which analysis can be conducted on the connections between variables. Local Capital Needs in OhioConversations with local government officials in Ohio have confirmed this connection between fiscal stress and delaying capital expenditures. Debbie Sutherland has served as the mayor of Bay Village, Ohio, since 2000. In good years, Sutherland said that the City would spend $700 thousand annually on road maintenance. Since 2008, spending on road maintenance has averaged only $100 thousand annually, meaning the City deferred several million dollars of road improvements. Sutherland offered an estimate of $3.5 million in deferred road maintenance and speculated about the number of good paying construction jobs this would have created in her jurisdiction. While some tax revenues have recovered since the recession, the cuts to state funding have kept capital investment low, pushing needed repairs further into the future (D. Sutherland, Personal Communication, April 15, 2016).Lakewood, Ohio, is seven miles east of Bay Village. Mayor Mike Summers described some of the challenges that his city faces in regard to capital investment. Lakewood is a 100-year-old community and the infrastructure is beginning to show its age. The City holds 52 thousand residents in 5.62 square miles, making it the most densely populated city in the Midwest outside of downtown Chicago. Lakewood faces the challenge of undoing and redoing 160 miles of public sewers in order to comply with stipulations of the Clean Water Act of 1972. Summers expects that Lakewood will need to spend about $200-400 million in order to reach compliance, a challenge in a community with 17 percent poverty and a median household income of $44 thousand. The city was already facing financial strain after the Great Recession and changes to the Local Government Fund certainly haven’t helped, Summers said. The longer these capital projects are deferred, the more expensive they become (M. Summers, Personal Communication, April 15, 2016). Bay Village and Lakewood are two examples of communities in Ohio where capital investment needs have gone unfulfilled, but they are certainly not the only examples. As the literature and the Fiscal Assessment Survey show, cutting capital expenditures is one of the first actions taken when a local government faces fiscal strain. Further research is needed to understand the extent to which this has occurred across the nation since the Great Recession. A deeper look at Ohio will be revealing to specifically understand how cuts to state revenue-sharing for local governments may further impact capital investment. Methods This study examines the ways in which changes to the Local Government Fund have affected revenue streams and capital investments in local governments in Ohio. A variety of analytical techniques are applied in order to answer these questions. First, descriptive analysis is conducted to understand the ways in which revenues to Ohio’s local governments have changed from 2007 to 2013. This analysis considers five separate streams of revenue: local government fund distributions, property tax, municipal income tax, charges for services, and fees, licenses and fines. Patterns of change in capital outlays are also examined across the cities over time. The measure used for this analysis is the statewide average per capita. Per capita measures are used here in order to normalize values across cities with different populations. Statewide average of percent of revenues for each revenue source is also examined, to better understand how much of cities’ total budgets are represented by each revenue source.Additional analysis was conducted on groups of cities based on population and median household income. The measure used for this analysis is the percent of total revenues from different revenue sources. This is used in order to determine whether cities tend to rely more heavily on different revenue sources based on their size or household income. This analysis is useful in understanding how cities may be differently impacted by the Recession and by changes to the Local Government Fund based on their reliance on different sources of revenue. However, because it is not relevant to the major research questions, these results are not included in the paper but instead in Appendices A and B, where they can be accessed by those interested in better understanding these differences. Regression analysis is used to determine whether certain revenue streams or different years had a significant impact on capital outlays. A pooled regression model is used, with observations by city and year. The dependent variable is natural log of capital outlays per capita. The per capita measure is used in order to remove population differences. Using the natural log allows elasticities to be interpreted from the results. The first set of independent variables reflects revenue collected per capita from the five sources discussed earlier: property tax, income tax, charges, fees, and Local Government Fund. These are also modeled using the natural log for the purpose of interpreting elasticities. Median household income, population density, and population are also included in the model in order to control for differences between communities that may contribute to their ability to make capital investments. Finally, year dummies are used for the years 2008 to 2013, in order to determine whether certain years showed a significant difference in capital outlays from the base year, 2007.The regression model used is as follows:ln(capital outlay per capita) = ?1(ln(property tax per capita)) + ?2(ln(income tax per capita)) + ?3(ln(charges per capita)) + ?4(ln(fees per capita)) + ?5(ln(LGF per capita)) + ?6(ln(median household income)) + ?7(ln(density)) + ?8(ln(population)) + ?9(Year=2008) + ?10(Year=2009) + ?11(Year=2010) + ?12(Year=2011) + ?13(Year=2012) + ?14(Year=2013) + constant + errorThis equation is modeled using a weighted least squares regression. The weighted variable is total revenue, measuring a city’s budget size. Weighting the model in this way gives greater weight to larger municipalities, producing more representative results. Robust standard errors are also used because heteroskedasticity (unequal variance) was found to be present in the model. The coefficients are interpreted at a 10% significance level.Two main hypotheses are tested in order to determine whether cuts to the Local Government Fund had a significant, negative effect on capital investment. First, it is expected that the coefficient (?5) for Local Government Fund revenues in a given year will be significant and positive. This would indicate that Local Government Fund revenues and capital outlays move together, indicating that spending on capital would decrease in the same year revenues from the Local Government Fund decrease. The null and alternative hypotheses are:Ho: ?5 = 0Ha: ?5 > 0Rejecting the null hypothesis here does not mean that Local Government Fund revenues went directly to capital spending. More likely, Local Government Fund distributions were entering cities’ general funds in order to provide other services. As the literature shows, if these funds were cut, it is likely that cities chose to redirect resources away from capital and towards current service obligations. Second, it is expected that the coefficients (?12, ?13, ?14) for the years 2011, 2012, and 2013 will be significant and negative. This would indicate that capital outlays were significantly lower in these years than in the base year, 2007. This would align with when cuts to the Local Government Fund were instituted, providing further evidence for the impact of these cuts on capital spending. The null and alternative hypotheses are:Ho: ?12 = ?13 = ?14 = 0Ha: ?12 < 0 or ?13 < 0 or ?14 < 0DataThe data used for this study reflect revenues and expenditures of Ohio cities from 2007 through 2013. All data are inflation-adjusted to 2013 dollars. This paper focuses only on cities in Ohio, leaving the other local government divisions of counties, villages, and townships for further study. This gives panel data of 250 cities in 7 years, with a total of 1,750 observations. Due to discrepancies in the cities reporting data each year, the regression model includes 1,225 observations, an average of 175 per year. Data were collected from the Ohio Auditor of State, Ohio Department of Taxation, and United States Census. All data was downloaded from the respective agency’s public website. Data from the Ohio Auditor of State reflect key revenues and expenditures reported by Ohio cities in their comprehensive annual financial reports. All cities reported in this data use generally accepted accounting principles. Data are collected on total governmental fund revenues and expenditures and on governmental fund revenues from property taxes, incomes taxes, charges for services, fees, licenses and fines, and capital outlays. All governmental funds are used because cities frequently use separate capital funds with certain revenue streams going directly to these funds rather than into a general revenue fund. Data from the Department of Taxation reflect distributions to cities from the Local Government Fund, both in direct distribution to the municipalities and in distribution from the county undivided funds. Municipal income tax rates are also collected from the Department of Taxation. A series of variables from the United States Census Bureau is added to these budget measures to understand whether responses may have varied in different kinds of cities. These include population, population density, and median household income. Descriptive statistics for the variables used in the models are shown in Figure 2 below.Figure 2: Descriptive Statistics (2007-2013, 250 Cities)(1)(2)(3)(4)(5)VARIABLESNMeanStandard DeviationMinimumMaximumCapital Outlays per Capita1,555164.7178.702,344Property Tax per Capita1,415146.5105.20963.1Income Tax per CapitaCharges for Services per Capita1,65379.0960.730461.1Fees, Licenses and Fines per Capita1,65450.0040.140314.3Local Government Fund Revenue per Capita1,58939.4123.261.344180.9Median Household Income1,66252,22223,09217,933207,069Density1,6622,3821,292313.09,222Population1,65628,96870,7484,348823,536Data source: Ohio Department of Taxation, Ohio Auditor of State, U.S. Census BureauResultsDescriptive AnalysisThe first set of results comes from descriptive analysis conducted on the data. First, statewide per capita averages on each of the variables were calculated for each year, in order to understand the general trends that have occurred in municipal finances across the state. Of course, there is no such thing as an average city as so much variation exists by size, household income, and funding structures. However, these averages help us more easily focus on the patterns of changes that have occurred over time across the state. Figure 3 shows the statewide averages for total revenue and expenditures per capita from 2007-2013, using inflation-adjusted 2013 dollars. As this chart shows, local governments across the state have seen, on average, a decline in the size of their budgets for governmental functions. In real dollars, statewide average total revenues per capita have fallen from $1,244 in 2007 to $1,127 in 2013, a 10% decline. Statewide average total expenditures per capita have fallen from $1,284 in 2007 to $1,154 in 2013, a 9% fall. In this data, total expenditures exceed total revenues due to the practices of budgetary basis accounting, where expenditures include spending as a result of debt issuance while revenues just show what was collected in that year.Data source: Ohio Auditor of StateRevenues and expenditures both hit a low in 2012, with average revenues per capita equaling $1,115 and average expenditures per capita at $1,143. This represents a 10% decline in revenues from 2007 to 2012 and an 11% decrease in expenditures. In 2013 revenues and expenditures appeared to see their first recovery since the Great Recession hit in 2008. However, this may be an eccentricity of the available data that requires further investigation. Revenues also appeared to recover slightly in 2010 after falling dramatically during the Great Recession in 2008 and 2009, whereas local government expenditures appear to have been on a steady decline.Looking at the different sources of revenue, it becomes clear that the declines in local government revenues overall can be pinpointed to a few key sources. Statewide averages for these different revenue sources as part of governmental fund revenues are shown in the following graphs, again using 2013 inflation-adjusted dollars. Income tax revenues are placed in a separate graph (Figure 4) due to a difference in scale, as income taxes represent more than half of revenues for most Ohio cities. Data source: Ohio Auditor of StateMunicipal income tax collection in Ohio shows a clear picture of recession and recovery in Figure 4 above. Income tax revenues fell dramatically in 2008 and 2009 as a result of falling wages and lost jobs during the Great Recession. Beginning in 2010, revenues began to increase and have done so slowly but steadily ever since. While economic recovery has certainly contributed to these increasing revenues, average municipal income tax rates have also been steadily increasing since 2007, shown in Figure 5 below. It would appear that local governments in Ohio have actively pursued income tax recovery through rate increases since the beginning of the Great Recession. Of the 250 cities in this dataset, 44 raised income tax rates from 2007 to 2013, with an average rate increase of 0.48 percentage points.Data source: Ohio Department of TaxationLooking at the other sources of revenue in Figure 6 below, it is clear that local governments have seen steadily declining property tax revenues, in real dollars. Average property tax revenues received by city governments have fallen from $172 per capita in 2007 to $132 per capita in 2013, a 23% decline. One reason why property tax revenues have not recovered from the Great Recession is because property tax reassessment does not occur every year. Therefore, the negative effects of a bad economy are often lagged in property tax collection. Data source: Ohio Auditor of StateRevenues from the Local Government Fund have also been decreasing since 2007, of course showing a more dramatic decline from 2011-2013 as a result of the aforementioned policy changes. Average revenue from Local Government Fund distributions fell from a high of $51 per capita in 2008 to $42 per capita in 2011. As the major cuts went into effect beginning in July 2011, the changes were realized by calendar years 2012 and 2013, which respectively showed $29 and $21 in average revenue per capita from this source. From 2008 to 2013, statewide average revenues to local governments from the Local Government Fund fell a full 58%.Revenue from charges for services and from fees, licenses, and fines show less change over the time period. Average revenue from charges for services was at its lowest in 2008 at $76 per capita in 2013 dollars, and increased to a high of $80 per capita in 2013, a 5% increase. Statewide average revenues from fees, licenses and fines fell consistently from a high of $53 per capita in 2007 to a low of $50 per capita in 2013, a 4% decline. Changes in revenue amounts collected from the previous two sources are difficult to connect directly to policy changes. In regard to charges for services, revenue collection can increase both due to increased rates and to increased utilization of the services. This is especially important to remember in this time period because demand for government services tends to increase when the economy is poor. For fees, licenses, and fines, a change in revenue could indicate a change in fee rates, in enforcement of the fees, or in utilization of activities that require fees or licenses. Again, it is hard to know which is responsible from this data, and further study will be needed in order to understand whether policy changes have occurred.Finally, Figure 7 below shows a steady decline in capital outlay in local governments across the state. From a statewide average of $202 per capita in 2007, average capital outlays declined to a low of $138 per capita in 2012, a 32% decrease. Capital outlay actually seems to have begun to recover in 2013, increasing to $156 per capita. However, this 2013 statewide average still represents a 23% decrease from 2007. While statewide data do not show a marked decline that occurred as a result of the changes to the Local Government Fund in 2011, these effects may be more apparent upon closer analysis. Data source: Ohio Auditor of StateThese statewide averages begin to paint a picture of the ways in which the Great Recession and changes to the Local Government Fund impacted local government finances. Overall, revenues and expenditures have declined. Revenues from income taxes have somewhat recovered, due in part to increasing municipal income tax rates. Revenues from property taxes and from the Local Government fund have fallen sharply, while revenues from charges and services have shown a slight increase and revenues from fees, licenses and fines have shown an even slighter decline. Finally, capital outlays have been steadily decreasing across the state from 2007 through 2008. A final piece of analysis was conducted on the percent of total revenues represented by each revenue source. This is a helpful piece of information in beginning to understand which revenue sources may have the greatest impact on expenditures. The results are shown in Figure 8 below. On average across the state, income tax revenues represent 45-50 percent of total revenues. This is distantly followed by property tax revenues, which tend to be around 12-15 percent of total revenues. Revenues from charges, fees, and the Local Government Fund all represent, on average, less than 10 percent of total revenue. The changes in percent over the years are not particularly relevant, because it does not necessarily indicate a policy change. If it is known that the percent represented by one source decreases, like the Local Government Fund, all of the other sources of revenue will necessarily increase in their percent, regardless of whether the local government increased collection from that source. Data source: Ohio Auditor of StateRegression AnalysisRegression analysis is used in order to further analyze how changes to local revenue sources, especially the Local Government Fund, have impacted capital investment. The details of the model are described in the methods section above. The results of the regression are shown in Figure 9 below. The R-squared for the model equals 0.218, which means that 21.8 percent of the variation in capital outlays is explained by the model. Analysis of the two hypotheses and other significant coefficients follows, as well as a discussion of the limitations of the model. Post-estimation diagnostics are included and discussed below. Coefficients are interpreted at 10 percent significance level. Figure 9: Regression Analysis – Weighted Least Squares by Total RevenueVARIABLESCapital Outlay per Capita (ln)Property Tax per Capita (ln)-0.109**(0.0540)Income Tax per Capita (ln)0.703***(0.0606)Charges per Capita (ln)0.0657*(0.0338)Fees per Capita (ln)-0.0380(0.0360)Local Government Fund per Capita (ln)-0.0458(0.0594)Median Household Income (ln)0.485***(0.103)Density (ln)-0.107*(0.0610)Population (ln)0.0692*(0.0385)Year = 20080.0208(0.112)Year = 2009-0.0636(0.110)Year = 2010-0.238**(0.111)Year = 2011-0.365***(0.109)Year = 2012-0.405***(0.112)Year = 2013-0.305***(0.117)Constant-4.055***(1.176)Observations1,225R-squared0.219Robust standard errors in parentheses*** p<0.01, ** p<0.05, * p<0.1Hypothesis 1The first hypothesis tested in the model was whether revenues from Local Government Fund distributions had a significant impact on capital outlays. It was expected that the coefficient for the Local Government Fund variable would be significant and positive, indicating that a decrease in Local Government Fund revenues would lead to a decrease in capital outlays in the same year, holding all other variables constant. Ho: ?5 = 0Ha: ?5 < 0The results here confirm the null hypothesis, as the coefficient on Local Government Fund revenues per capita (?5) is not significant. This could be due to a model specification error. It is possible that the variable would be significant if the dependent variable were lagged by one year. This would indicate that when Local Government Fund revenues fall in one year, a city would decrease their capital investment in the following year. Modeling in this way could be a possibility for future research. Hypothesis 2The second hypothesis tested was whether capital outlays were significantly lower in 2011, 2012, and 2013 than in 2007, as these years align with when the cuts to the Local Government Fund went into effect. Ho: ?12 = ?13 = ?14 = 0Ha: ?12 < 0 or ?13 < 0 or ?14 < 0These coefficients were all found to be significant and negative, allowing for rejection of the null hypothesis. This means that capital outlays per capita were significantly lower in years 2011, 2012 and 2013 than in the base year of 2007. The coefficient for the year 2010 is also significant indicating that capital outlays per capita were also significantly lower in this year than in 2007. In order to interpret the results for this hypothesis, the coefficients need to be converted using the formula (e?-1)*100. This will give a percent value that can be interpreted as the difference between the dependent variable in the base year and in the year indicated by the coefficient. In 2010, capital outlays per capita were 21.2 percent lower than in 2007. This difference increased in the following years, with capital outlays per capita 30.1 percent lower in 2011, 33.3 percent lower in 2012, and 26.2 percent lower in 2013. With cuts to the Local Government Fund having been enacted in mid-2010 and effective beginning in mid-2011, these findings offer compelling evidence that local governments in Ohio have reduced spending on capital outlays in response to cuts from the local government fund. There are certainly valid alternative explanations for why capital outlays were significantly lower in these years than in 2007. It is possible that the effects of the recession were delayed for local governments property taxes, the second largest source of revenues for local governments, because these are awarded in arrears and reassessed in three year cycles. This means that although property values began to fall with the recession in 2008 and 2009, this did not reach local government budgets until the following year. This is an especially compelling alternative explanation given that revenues were significantly lower in 2010 as well as 2011 through 2013. This could also be due to the nature of budget cycles, where spending for 2008 and 2009 was likely set before the onset of the Great Recession. Further research on other expenditure measures could reveal whether cuts to other expenditures happened at the same time as for capital outlays. Revenue VariablesThree revenue variables were found to have significant coefficients: property tax per capita, income tax per capita, and charges per capita. The coefficients for income tax per capita and charges per capita are both significant and positive, indicating that capital outlay moves in the same direction as these revenues. For every 1 percent increase in income tax revenue per capita, capital outlays per capita increase by 0.703 percent, holding other variables constant. The coefficient for charges for services indicates that for every 1 percent increase in revenue from charges per capita, capital outlays per capita increase by 0.0657 percent, holding all other variables constant. This shows that the relationship to capital is stronger for income tax revenues than for revenues from charges. This is not surprising given the high percent of revenues represented by income tax revenues for local governments in Ohio. It is possible that the coefficient for income tax revenue would be significant and positive for any expenditure category due to its outsized influence on the budget as a whole. This result could have further implications for cities that are considering ways to increase their revenues that can be used for capital spending. This will be discussed further in the conclusions section. The coefficient for property tax revenues per capita is also significant, but negative, which is surprising. The value indicates that a 1 percent increase in revenue from property taxes leads to a 0.109 percent decrease in capital outlays per capita, holding all other variables constant. It is possible that this is capturing differences between cities. The descriptive analysis in Appendices A and B shows that smaller cities tend to collect more of their total revenues from property taxes than do larger cities. This property tax coefficient then may be showing that smaller cities have lower capital outlays per capita than larger cities, which is confirmed by the coefficient on population discussed below.Control VariablesThe coefficients for the three control variables included in the model were also found to be significant. The coefficient for median household income indicates that a 1 percent increase in median household income in a community can lead to a 0.485 percent increase in capital outlays per capita, holding other variables constant. The significance and direction of this relationship is not surprising, given that a higher median household income creates a higher tax base, increasing revenues and likely spending on all services.The coefficient for density is negative, indicating that a 1 percent increase in households per square mile is associated with 0.107 percent lower capital outlays per capita. The coefficient for population is positive, meaning that a 1 percent increase in population is associated with a 0.0692 percent increase in capital outlays per capita. This difference in direction between density and population in their relationship to capital outlays may be capturing the unique challenge of inner-ring suburbs like Lakewood. Generally, larger cities may be better off with their capital spending than smaller cities, but mid-sized cities with high population density may actually be worse off than small cities with low population density. This makes sense given that inner ring suburbs also tend to have a lower median household income, which the model confirms also has a negative impact on capital spending. Post-Estimation DiagnosticsFigure 10 below shows the correlation matrix for the variables included in the regression analysis. The only correlation that is high enough to create a concern of multicollinearity is that between population and total revenue. Thus, these variables are not both included in the regression analysis; instead, population is included in the regression while total revenue is used as the weighting variable for weighted least squares analysis. Figure 10: Correlation MatrixInitial regression analysis was run without the use of robust standard errors. This first regression was then tested for heteroskedasticity using the Breusch-Pagan/Cook-Weisberg test. The null hypothesis for this test is constant variance. The test returned a chi-squared value of 34.82, with a probability of 0.000. This provides evidence that heteroskedasticity is present in the model. Robust standard errors are used in the final regression model in order to mitigate the effects of heteroskedasticity on the model.After the final regression was run with robust standard errors, variance inflation factors (VIF) were calculated in order to ensure that multicollinearity was not a problem in the model. The decision rule here is that multicollinearity is present if a VIF for a variable is greater than 2.50. The results of the VIF calculation for the model are shown in below. Because none are greater than 2.50, multicollinearity is not a concern. This confirms the assumptions made using the correlation matrix prior to the regression. Figure 11: Variance Inflation FactorsLimitationsThere are limitations to the regression model that need to be considered in interpreting the results. Most important are the limitations present in the data. The regression model contains 1,225 observations out of a grand total of 1,750 observations. This is because there were cities that did not report capital outlays and cities that incorrectly reported property and income tax revenues. All city/years to which this applied were removed from the analysis. The data are also unaudited, so there is a risk that some of the data are inaccurate. Staff members at the Auditor of State’s office expressed particular concern about the accuracies of the data from the smallest cities. In practice, many errors were found in the data from the six large cities, which were corrected by referencing the Comprehensive Annual Financial Reports directly from those cities. Knowing that these errors exist does cast concern about further data inaccuracies. Omitted variable bias is certainly present in the model, which could be the reason for a low R-squared value. There are many determinants of capital outlays that are not contained within the model, including age of infrastructure, federal funding, and some measure of capital needs. Further research should determine ways to effectively measure this need and include this in the regression model. There are also many other reasons why the variables for the years 2010-2013 may be significant that are also not included in the model. Other measures of economic health could be added in future iterations in order to more effectively remove some of the variation between years and focus more fully on the Local Government Fund. ConclusionsThe findings of this research offer evidence that capital outlays significantly decreased in the years following the Great Recession. Compared to 2007, capital outlays were 21.2 percent lower in 2010, 30.1 percent lower in 2011, 33.3 percent lower in 2012, and 26.2 percent lower in 2013. Further research is needed to more fully understand the connection between changes to the Local Government Fund and capital investment. Revenues from income taxes and charges for services are positively related to capital outlay per capita, which is not surprising especially given the outsized impact that income tax revenues have on municipal government budgets in Ohio. Median household income and population were also positively related to capital spending per capita, while a negative relationship existed between capital spending and population density.Policy ImplicationsLocal government leaders are interested in the results of this research as it relates to alternatives for funding from the state government. Of course, local government leaders would like to see the Local Government Fund returned to former funding levels, but this is highly unlikely given the state of politics in Ohio. The administration and legislature are largely the same at the moment as they were when the Local Government Fund was reduced, and these same politicians cannot be expected to reverse their previous action. Instead, it is important for those lobbying on behalf of local governments to emphasize the importance of the remaining half of the Local Government Fund in order to prevent further reduction.Additionally, some Ohio mayors have begun to discuss the possibility of a state-run capital distribution program for local governments. This would replace some of the funds lost from the Local Government Fund and apply them directly to local capital investment. Bay Village Mayor Debbie Sutherland and Lakewood Mayor Mike Summers offered further details of how this type of policy would be structured. There are already competitive grant programs available at the state and federal level for capital projects, but Sutherland and Summers would like to see something different here. Competitive grant programs often require matching funds, leaving communities who need the most help unable to obtain any of the funds. They would like to see a revenue-sharing program with money that is automatically distributed to municipalities based on a funding formula, with no specific project requirement or application. Summers proposed that the formula for the Local Government Fund be replicated for this new infrastructure fund, as this formula has been developed and honed over the course of many decades. It is well respected across the state and takes into consideration variables including population size, age of city, and poverty level (D. Sutherland and M. Summers, Personal Communication, April 15, 2016).Both Sutherland and Summers emphasized the importance of restricting the use of these funds to physical infrastructure work, including redesign, improvements, and new construction. Summers emphasized the ubiquity of declining capital spending across the state, saying we “all have plenty of other issues, but we all have these issues.” Sutherland emphasized the political feasibility of tying these funds directly to capital. Cutting the Local Government Fund was intended to restrain local government spending, she said. The legislature is not interested in hearing about cuts to police and fire or anything personnel related, because this was part of the intent of the original policy. They are interested in the impact on infrastructure, because state legislators understand the link between local government investments in capital and economic development across the state (D. Sutherland and M. Summers, Personal Communication, April 15, 2016).Directions for Future ResearchFuture research can deepen the findings of this study in a number of ways. Similar research can be done using a more comprehensive dataset that includes all cities, villages, counties, and townships in Ohio. All of these local governments were affected by the changes in different ways, and it will be important going forward to get a fuller picture of the experiences of each. Case study analysis could be beneficial to better understand how cutting the Local Government Fund has specifically impacted capital spending. Future study should also include more measures of capital need like age of infrastructure, miles of road requiring improvements, and other of these more physical measures. This would be beneficial in understanding not just what local governments have spent on capital investment but also what they have not spent.Continued research is also needed on other effects of cutting the Local Government Fund. The Fiscal Assessment Survey in 2013 revealed many other impacts of these policy changes on revenues and expenditures for local governments, including reductions in staffing and other services. Continued quantitative research will be beneficial in clarifying and supplementing these other findings of the survey. Ohio’s state and local government leaders should continue to invest in research of these impacts in order to inform municipal finance policy into the future. The Great Recession had a major impact on the fiscal health of state and local governments in Ohio. The decision of the state government to cut the Local Government Fund had lasting impacts on the spending decisions of local governments across the state. Academic literature, survey research, and anecdotal evidence all point to cuts in capital investment as an impact of these state policy changes. While this study finds preliminary evidence that capital outlays declined after the Great Recession, further research is needed to solidify the relationship between capital spending and distributions from the Local Government Fund. BibliographyBegala, J. (2010, June). Thinking the Unthinkable: Finding Common Ground for Resolving Ohio's Fiscal Crisis. Center for Community Solutions.Hoene, C., & Pagano, M. (2003). Fend-for-yourself Federalism: The impact of federal and state deficits on America's cities. Government Finance Review.Jimenez, B.S. (2009). Fiscal Stress and the Allocation of Expenditure Responsibilities between State and Local Governments: An Exploratory Study. State and Local Government Review, 41(2), 81-94. Marshall, A. (2011, May 29). Ohio’s $8 billion budget hole: Was it really that big? Cleveland Plain Dealer. Retrieved from Martin, L., Levey, R., & Cawley, J. (2012). The "New Normal" for Local Government. State and Local Government Review , 44 (1), 17-28.Mullins, D., & Pagano, M. (2005). Local Budgeting and Finance: 25 Years of Developments. Public Budgeting and Finance.Oates, W. (1999). An Essay on Fiscal Federalism. Journal of Economic Literature, 37 (3), 1120-1149.City of Upper Arlington. (2013). Ohio Municipality Fiscal Assessment Survey. Retrieved from .Ohio Department of Taxation. (2011). Changes to the Local Government Fund, as enacted by FY12-13 state operating budget. Retrieved from HYPERLINK "" tax.. Ohio Department of Taxation. (2013). Changes to the Local Government Fund and the Public Library Fund, as enacted by the FY14-15 state operating budget. Retrieved from tax.. Ohio Department of Taxation. (2005). Ohio's Local Government Funds. Retrieved from tax.. Ohio Department of Taxation. (2016). Tangible Personal Property Tax. Retrieved from tax.. Ohio Department of Taxation. (2007-2013). LG5: County Undivided Local Government Funds - Amounts Distributed within Counties by County Budget Commissions by Subdivision or Subdivision Class (LG5). Retrieved from tax.. Ohio Department of Taxation (2007-2013). State & Local Government Fund - State Distribution to Counties & Municipalities (LG1). Retrieved from tax.. Ohio Department of Taxation (2007-2013). Municipal Income Tax-Tax Rates and Amounts Collected, by Municipality?(LG11). Retrieved from tax.. Ohio Auditor of State. (2007-2013). Summarized Annual Financial Reports, Cities. Retrieved from . Scorsone, E., & Plerhoples, C. (2010). Fiscal Stress and Cutback Management Amongst State and Local Governments: What have we learned and what remains to be learned? State and Local Government Review , 42 (3), 176-187.Testa, J. W. (2011). Ohio Department of Taxation 2010 Annual Report. Retrieved from tax.. Appendix ARevenue Breakdown by PopulationThe next set of results looks at the cities divided by population in order to better understand how policy changes may have impacted cities of different sizes in unique ways. The cities are broken down into five groups roughly equal groups based on population: Less than 10,000, 10,000-19,999, 20,000-39,999, 40,000-99,999, and greater than 100,000. These divisions represent a roughly equal division of the number of cities in each group, with the exception of the six largest cities (Columbus, Cincinnati, Cleveland, Dayton, Akron, and Toledo), that occupy the greater than 100,000 group.The measure used in this portion of the analysis is percent of total governmental fund revenues. This is used in order to determine where differences may exist in the revenue make-up decisions between cities of different population sizes. This information can help in understanding how policy changes to one of these revenue streams may impact cities in different ways based on their populations. However, the problem with this measure is that it is difficult to compare across time and across different revenue streams. This is because if one type of revenue decreases, as the Local Government Fund did, the other types of revenue will increase as a percentage of total revenues, even if the amount collected did not change. Thus, an increase or decrease in the percent of total revenues collected from a certain sources may not indicate that a policy change has occurred. Figure 12 below shows the difference in Local Government Fund distributions as a percentage of total revenue. This can offer some understanding of how these policy changes may have affected cities of different sizes in different ways. Interestingly, cities with populations of greater than 100,000 and of 40,000-99,999 tended to have a higher percentage of revenues covered by Local Government fund distributions, followed by the smallest cities with populations less than 10,000. Small and mid-sized cities with populations from 10,000-39,999 have shown less reliance on local government fund revenues than larger or smaller cities. Cities of all sizes saw the results of declining local government fund distributions in 2012 and 2013. Data source: Ohio Department of Taxation, United States Census BureauFigure 13 shows the difference in property tax revenues as a percent of total revenue, divided by population. Notable here is the difference between the large cities, represented by the light blue line, and all of the smaller cities. Larger cities tend to rely more heavily for revenue from income taxes, as taxes are typically collected by the city of employment, not residence. The decrease in reliance on property tax revenues is also clearest in these large cities and in the second most populous group of cities (40,000-99,999). Cities under 10,000 people rely more heavily on property tax collection as a percentage of revenue than cities of other sizes. Often, large cities have more industrial properties, which reduces the need to rely on property taxes collected from individual taxpayers. Data source: Ohio Auditor of State, United States Census BureauIncome tax revenues in Figure 14 appear to tell the complimentary story. It is first important to note that income taxes represent about 40-50 percent of total governmental fund revenues for most municipalities across the state. This shows that municipalities in Ohio face different financing decisions than many states in the country, because nearly half of revenue comes from a source to which many local governments across the country do not have access. Secondly, whereas property taxes have been showing a decline as a percentage of revenue, income taxes show an upward trend. This follows the findings above that the statewide average municipal income tax has increased since the Great Recession. While this may not represent a widespread policy change to greater income tax collection, it can be said that a larger percentage of revenues are being represented by income tax collection.Data source: Ohio Auditor of State, United States Census BureauThe distribution between cities of different sizes is also not surprising, given the findings for property tax revenues. Income taxes represent a larger percentage of revenues for major cities than for the smallest cities with populations under 10,000. However, it is interesting to note that the 10,000-19,999 population group appears to be the most reliant on income taxes, and saw a sharp spike in their reliance during the recession in 2008. This may be due to a larger hit to property taxes in these small cities. Additionally, mid-sized cities with populations from 40,000-99,999 are also the least reliant on income taxes. This could be because many cities this size are located in suburban areas where most of the jobs are located in the urban core, limiting the ability of these cities to raise sufficient revenue from their income tax base. These could also be some of Ohio’s older manufacturing cities that are seeing a declining job base and thus are less likely to depend on income tax revenues. Lastly, the Figures 15 and 16 below show the percent of total governmental fund revenues represented by charges for services and by fees, licenses and fines. The connection between these revenues and population is not readily apparent. Small cities with populations of 10,000-19,999 appear to have the lowest percentage of revenue represented by charges for services, whereas cites under 10,000 tend to have a higher percentage represented by these sources. The largest and smallest cities showed a lesser reliance on fees, licenses, and fines than any mid-sized cities. The percentages of revenue represented by charges for services appears to have increased since 2008 for most cities, whereas there is not a consistent trend for fees, licenses and fines. Data source: Ohio Auditor of State, United States Census BureauData source: Ohio Auditor of State, United States Census BureauAppendix BRevenue Breakdown by Household IncomeThe following set of descriptive results examines how changes have varied in groups of cities based on 2013 median household income. For this analysis, the six largest cities in the state are removed from the analysis, as their ability to respond to fiscal changes varies significantly from smaller cities that may have a similar median household income. The remaining 244 cities were divided into nearly equal groups representing Low, Middle, and High Income cities. The groups are listed below with the range covered and number of cities in each. Group Label2013 Median Household IncomeNumber of CitiesLow Income$17,933-40,08082Middle Income$40,219-54,22581High Income$55,942-207,06981Figure 17 below examines the differences in percent of revenues represented by Local Government Fund distributions by income group. This graph shows that cities with High median household income have relied less on Local Government Fund distributions as a source of revenue than cities with lower household incomes. This makes sense in considering the reasons why a Local Government Fund was created in the first place. Distributing state tax revenues from the state to the local level sought to even some of the inequalities that exist when local governments rely primarily on property tax revenues. Data source: Ohio Department of Taxation, United States Census BureauProperty tax as a percent of revenue shows the clearest difference among cities by household income, as shown in Figure 18. Higher income cities rely more heavily on property tax revenues than lower income cities. This makes sense, as it can be presumed that cities with a higher median household income also tend to have higher property values that garner greater revenues from this source for the city government. With higher property values, more revenue can be collected at a lower rate. Data source: Ohio Auditor of State, United States Census BureauIncome tax as a percent of revenues did not tell a clear story when divided by median household income in Figure 19. It appears that while low-income cities tended to rely more heavily on income taxes than middle and high-income cities prior to the great recession, middle and high-income cities increased their reliance on income tax by a greater amount post-recession and after the changes to the Local Government Fund. Again, this does not necessarily indicate that a policy change has taken place. Cities with higher incomes may have seen incomes recover more quickly and more fully after the recession than cities with lower incomes. High-income cities may have also had more ability to raise income tax rates than low-income cities in the wake of falling income tax revenues. This will require further analysis on income tax rates to be determined.Data source: Ohio Auditor of State, United States Census BureauCharges for services and fees, licenses, and fines in Figures 20 and 21 show a more consistent pattern by income division than by population division. High-income cities have been consistently less dependent on charges for services than low- and middle-income cities. Low-income cities have consistently been more dependent on fees, licenses and fines than middle or high-income cities. This follows with the apparent tradeoff between these revenue sources as compared to income and property tax revenues. When these tax bases are larger, cities are more likely to draw from them than to pull from these other revenue sources. Data source: Ohio Auditor of State, United States Census BureauData source: Ohio Auditor of State, United States Census Bureau ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download