Money or Spending? Why Increased Finances Do Not Lead to Better Social Outcomes

Public expenditures are powerful tools of government to stimulate, direct and most crucially, democratize development; ensuring that the least of society does not get left behind. When allocated properly, public expenditures provide for the poor what markets easily overlook, such as education, health and other social services (Keefer and Khemani 2003). In 1991, the 8^th Congress of the Philippines enacted the Local Government Code (LGC) that devolved significant powers and responsibilities to local government units, or LGUs (Atienza 2006). None of these powers and responsibilities were arguably more crucial than the increased autonomy in fiscal policy—both in the source of finances and decision making in public spending (Hutchcroft 2012). The logic being that as LGUs are both spatially and knowledgeably closer to the people, they are the most efficient and effective institutions of governance to ensure development reaches the grassroots (Asian Development Bank 2005).

            However, what if public expenditures are not spent prudently and are largely misplaced?  Indeed, in the Philippines, this has been the case. Increased finances and decision making powers provided by the LGC does not translate to any significant increase in social services spending, even when poverty and population continues to increase (Diokno 2012). The failure of LGUs to fulfill the promises of decentralization have real and terrible impacts on the people at the ground. Instead of ensuring equitable development, these failed institutions merely exacerbate the vicious cycle of poverty.

The study provides counterfactual evidence to the common hypothesis that increased local finances and autonomy in public expenditure decision making does lead to better social outcomes in the form of increased social services spending. Whereas previous studies merely analyze statistics through convenient comparison at face value, the study contributes by employing a regression analysis of collated data of 80 Philippine provinces from 2009-2013. It finds that social services spending is either unaffected, barely increased, or even decreased by an increase in the internal revenue allotment, local revenues, presence of civil society organisations, poverty and population when increases in all are expected to follow an increase social services spending (Capuno 2005; Diokno 2012; Llanto 2012; Yilmaz et al 2008). Ultimately, the study finds that it is the absence of effective social accountability mechanisms and failed institutional design that are largely unable to check patronage and corruption; which, in turn, incentives the bad spending behaviour of LGUs in the Philippines. Further studies might want to explore other proxies for accountability and develop the accuracy of the dataset.

Review of Related Literature

Local Government Finances—The IRA and Local Revenues

            The intergovernmental transfer system in the LGC of 1991 (Title III, Chapter One), the Internal Revenue Allotment (IRA), provided for significantly higher, predictable and automatically-released funds for LGUs (Diokno 2012). It increased the share of LGUs from 20% of all internal taxes to 40% and was now mandatory rather than under the discretion of the national executive. The amount of the allocation followed a formula; provinces and cities received a share of 23% each, municipalities received 34% while barangays received 20%. These were all formulated with respect to population (50% weight), land area (25%) and equal sharing (20%). The IRA remained to be the central transfer scheme, but local government units also received shares from national wealth derived within their respective areas and wider tax powers (Llanto 2012).  Yilmaz et al. (2008, p. 22) argues that “the design of intergovernmental transfer systems has implications for accountability because it affects the fiscal dependency on central government and local revenue-raising ability.” Indeed, the IRA, as the chief intergovernmental transfer system, has unexpectedly had negative outcomes and influences on LGUs’ behaviour.

            The higher, predictable and formula-based grant system has resulted in greater dependence of LGUs on the IRA. Instead of being stimulative or motivating LGUs to supplement the IRA with local revenues for a more complete fiscal capacity, the IRA has increasingly become a substitute for local revenue generation (Diokno 2012; Llanto 2012). According to the 2008 data of the Bureau of Local Government Finance, the IRA accounted for 78.5 percent of total revenues for municipalities, 43.3 percent for cities and 73.6 percent for provinces. As the IRA’s share in the total LGU income increased through the years, tax revenue and nontax revenue shares have decreased (Llanto, 2012). There are political and economic explanations for such a trend. Politically, it is only rational for local politicians to rely on the IRA instead of taxing his constituents which might breed unpopularity and have electoral consequences (Hutchroft, 2012). Economically, the dependence on IRA is a result of the absence of a tax base for local revenue generation in most LGUs (Medella, 2012).

            Aside from LGU dependence and lag in local revenue generation, the local fiscal capacity of LGUs is also lacking. As mentioned before, most LGUs simply do not have wide enough of a tax base to generate revenue. According to World Bank data in 1994, LGUs with relatively more narrow tax bases such as provinces and municipalities, when compared to cities, receive 46 percent and 47 percent of the cost of devolved functions, respectively. Further, they divide the pie of the IRA with more players when compared to cities (Llanto, 2012).

            Exacerbating the problem of their lack in resources, LGUs have also displayed patterns of spending that indicate misplaced priorities. According to data from the Bureau of Local Government Finance in Table 1, the percentage of general public services in the local budget rose from 40.5 percent in 2001 to 44.1 percent in 2008; this portion of the budget is used for general administrative expenses, which are often subverted for patronage uses (Hutchcroft 2012; Llanto 2012). The budget share of that under other purposes also increased from 12.4 in 2001 to 17.2 in 2008, suggesting an overall opaqueness to the budget process as contrary to transparency. Surprisingly, the budget share of social services fell from 26.2 percent in 2001 to 20.3 percent in 2008. Economic services also fell from 18.6 percent to 15.1 percent, in 2001 and 2008 respectively. According to former budget secretary and economist Benjamin Diokno (2012), “these expenditure categories are expected to grow with population growth. The fact that their budget shares fell in the face of rising population (and poverty incidence) suggests deterioration in the provision of social services at local levels.”

Decentralization and the Accountability Deficit

            The consequence of the failure fiscal decentralization to ensure both capacity and effective spending is the imbalance of development across and within regions. According to Capuno (2005), what is more critical to regional growth is not increased finances and devolved functions of LGUs but the prudent exercise of such powers and responsibilities. Decentralization reforms have often been introduced without consideration of the implications of accountability; it places emphasis on increasing discretionary powers of LGUs on the logic of efficiency and effectivity (Yilmaz et al. 2008). Indeed, in the Philippines, both scholars and government officials have took notice of the wide discretionary space local chief executives exercise, often misused, powers in. Secretary Jude Esguerra (2001) of the National Anti-Poverty Commission observe that mayors are “budget dictators” who holds the power of the purse over councilors. He further notes that coupled with high proportion of local spending in personnel, mayors utilize a spoils system where patronage is the basis of who he hires within the city hall, with complete disregard for civil service laws. Paul Hutchcroft (2012) notes that resources are often spent on projects that provide the most visible impact such as basketball courts with backboards declaring the politicians name or the welcome signs found at boundaries. The allocation of resources are “likely to be based more on electoral considerations than technical assessments” (Hutchcroft 2012). Benjamin Diokno (2012) provides empirical evidence that LGUs spend more on patronage-rich and opaque items of the budget, which are general public services and other purposes respectively. Finally, Joseph Capuno (2005) argues that the low quality of governance, local public services and uneven regional development is because of the poor compliance of LGUs to local consultative bodies and other governance features of the LGC. Patronage, also, is largely expected and demanded of local chief executives. He argues, “where there is wide people’s participation and competent leadership, the levels and quality of local public services have been noted to improve” (Capuno 2005, p. 3).

            These different observations on the wide discretionary space of LGUs are paralleled with a diverse set of proposed solutions. The existing literature generally agrees that elections cannot be relied upon as effective mechanisms to sanction misbehaving local chief executives (Capuno 2005; Diokno 2012; Hutchcroft 2012; Keefer and Khemani 2003; Yilmaz et al. 2008). Keefer and Khemani (2003) particularly single out on the imperfections of the political market that disrupts the translation of popular preferences into government policy; particularly in the lack of information. They in turn advocate for more transparency, availability and the expanded reach of information to bridge this gap. Capuno (2005) argues that because existing electoral laws and procedures are weak in the Philippines, the participation of the people through civil society organizations are better able to exact welfare benefits and increase the quality of social services. Yilmaz el al. (2008) agrees with Capuno that because the internal control environment and internal audits are non-existent in the Philippines, accountability exacted from civil society organizations are the last and best line of defense against misbehaviour. In particular, the authors argue these citizen organizations must participate in both the budgeting and expenditure process. Finally, Hutchcroft (2012) dismisses the misplaced hope placed on local reform strategies. He argues that the pervasive patronage structure in the Philippines operates throughout local-national dynamics that any hope of change should be directed at changing the institutional design of elections to be more proportional to successfully curb patronage in the country.

       
     The paper takes off from the ideas and observations made by these scholars and attempts to empirically and quantitatively test the validity of their arguments. Because the existing literature is largely in narratives and small-n cases, they necessarily sacrifice accuracy for depth. The value of this study lies in its ability to provide more comprehensive and accurate evidence to fill the gap of existing qualitative studies. By providing a statistical and large-n evidence, the study lays a stronger foundation for the existing hypothesis and proposed solutions.

Methodology and Operational Definitions

            Using a Philippine provincial level dataset recorded across the years 2009-2013, the study employs a panel-data estimation technique. We assume that the dependent variable Y, which indicates social services spending is linked to the amount of internal revenue allotment and a host of other possibly explanatory variables as follows:

            The model is in a log-level functional form as it is best able to capture the findings of the research in interpretation and generates the highest explanatory value. The absence of any interaction of variables and quadratic variables are consistent with both theory, empirical observations, tests against different models (See Table 2).

            The dependent variable of social services spending is used as a metric for social outcomes and welfare. Public expenditure on social services have the most direct impact on the welfare of the poor as they provide a steady and reliable stream of goods that are directly consumed, protecting the poor from adverse market effects (Snyder and Yackovlev 2000).

            The independent variable of the IRA is used as both a metric for LGU finances and patronage, representing the lack of accountability in public finances. Hutchcroft (2012, 113) argues that “for all the celebrated talk of promoting local autonomy and instituting fiscal decentralization, the IRA is also very much a story about the enhanced access of local politicians to patronage resources.” The hypothesis is that an increase in finances, in the form of the IRA, translates to more social services spending, but only to a marginal level as it is still largely misused for patronage purposes.

            The other independent variable of local revenues is used both as a metric for LGU finances and degree of autonomy. Whereas the LGC of 1991 provides LGUs with a more expansive power of taxation and collection of other fees to generate revenue, the share of local revenue from the total revenues of LGUs have been on a decline (Diokno 2012; Llanto 2012). This suggests that far from the goal of autonomy, LGUs are still largely dependent on national government for its resources. The hypothesis is that an increase in finances, in the form of local revenues, increases social services spending but less than the increasing effect of the IRA. This hypothesis reflects the small share of local revenues on total revenues and the dependency on the IRA.

            Another independent variable are civil society organizations, used as a metric for accountability. Different scholars have noted of the failure of formal legal-institutional accountability mechanisms and look to societal accountability as alternatives to check on government (Arugay 2005; Capuno 2005; Yilmaz et al. 2008). More than legal-institutional mechanisms, the role of civil society organizations as watchdogs are more political in nature and thus, a variable to the equation. The hypothesis is that an increase in the number of civil society organizations translates to more social services spending as LGUs are checked to spend more prudently.

            The final two independent variables are poverty, as measured by poverty incidence or the number of households under the poverty line, and population. It is well established that the logic of public expenditures, on social services most especially, is to provide for the poor what markets fail to (Keefer and Khemani 2003; Snyder and Yackovlev 2000). It is therefore expected that as poverty incidence increases, there ought to be more social services spending. Also, population growth is support to coincide with increased spending on social services for the quality to not degrade (Diokno 2012).

 Data

To find the effect of the IRA on social services spending, the study employs a collated a provincial-level dataset obtained from the Bureau of Local Government Finance, Local Government Academy, Caucus of Development NGO Networks (CODE-NGO), and the National Coordinating Statistical Board. Specifically, the dataset constitutes a panel of 80 provinces from 2009 to 2013. However, data for the variables poverty and population are taken from the years 2011 and 2010, respectively, as these years are the latest of censuses that are conducted in intervals and are assumed to be constant. The data for NGOs from 2011 are also assumed to be constant due to the difficulty of obtaining yearly data. The list of provinces excludes Dinagat Islands due to the absence of significant data throughout the period covered. Table 2 shows the descriptive statistics of the regression variables used.

         
   The variables of social services spending, IRA, and local revenues were made uniform to be measured in millions. The variables of NGO and population are to be taken as nominal. Finally, the variable of poverty is to be taken in percentages. Under the minimum columns of social services, NGO and poverty, there are noticeable zeroes. First, Sulu in 2011 lacked official data on social services spending. Second, there were unavailability of NGO data in the provinces of Basilan, Sulu and Tawi-Tawi. Finally, Batanes posted zero percent poverty incidence in the year 2011.

Results

            Table 3 presents the regression table of the effect of the IRA and other important control variables on social services spending. The estimates clearly suggest evidence in favour of the first hypothesis. It shows for every one million peso increase in IRA, social services spending increases by only 0.0182% or 182 pesos. The very minimal increase of social services spending confirm that a large chunk of the IRA is spent on other items. Further data from Diokno (2012) in Table 1 provide evidence for another assertion of the first hypothesis, that large parts of the IRA and LGU finances in general are misused for patronage and not transparent purposes.

            The second hypothesis, however, must be completely rejected. The regression results show that local revenues have no statistically significant effect in terms of increasing spending on social services. But instead of completely disposing of the logic of the hypothesis, the results merely validates the argument on a higher level. Whereas the hypothesis expected a positive but less effect than the IRA to signify dependence, the insignificance of the variable strengthens the assertion that LGUs are extremely dependent on the IRA in terms of financing. The evidence shows that the goal of the LGC of 1991 to provide LGUs “genuine and meaningful local autonomy” has failed.

             The results on NGOs is an interesting case. Whereas most if not all literature argue that increased people’s participation in the form of NGOs increases both the level and quality of social services (Bräutigam 2004; Capuno 2005; Krafchik 1996; Yilmaz et al. 2008), the estimation shows otherwise: for every increase in the number of NGOs, social services spending decreases by 1.23% or 12, 300 pesos. It is important to note that the literatures noted above were qualitative studies that did not employ statistical analysis, thus may explain the difference in outcomes. It is also possible that the dataset employed in this study is inaccurate as it is but the number of surveyed NGOs for the profiling purposes of the Local Government Academy and may not reflect the actual number of NGOs in the province. These considerations aside, the negative effect of NGOs on social services can be explained through a pluralist model of democracy. NGOs can be likened to interest groups that pressure the state for different policies and allocation preferences. It is important to understand that NGOs do not have common agendas and may be interested in allocations outside of social services spending. NGOs, therefore, may be rechanneling public expenditure away from social services and into their respective preferences.

Conclusion

Governments are powerful actors with the necessary tools in public expenditures to combat market failures that often hit the poor the hardest. The Local Government Code of 1991 sought to make public expenditures more efficient and effective by providing local government units with extensive autonomous powers and capacities. The logic being that as LGUs are both knowledgeably and spatially closer to the people, they are most effective in not only achieving development but more importantly, an equitable one. To achieve this, public expenditures made by LGUs must be based on the standpoint of what is optimal for development. The empirical evidence, however, points otherwise.

 The significant increase in resources provided by the IRA does not translate to significant increases in social services spending, even when the average poverty incidence in all provinces is at 26 percent. The same logic follows as poverty and population are deemed to be statistically insignificant in determining the level of social services spending. Further, local revenues, made higher by the IRA’s provision of more expansive taxing powers, is also statistically insignificant, suggesting that LGUs are still largely dependent on national government for resources. Finally, the decreasing effect of NGOs suggests that government may be privileging certain interests at the expense of development. NGOs, in effect, are failed experiments of social accountability.

To further the research, scholars may seek to include other variables for both internal and social mechanisms of accountability. A more accurate dataset on NGOs is also warranted and may change the outcome of this study.

Despite all considerations, the evidence leads the study to conclude that the current project of decentralization in the Philippines has failed. Despite the significant transfer of resources, responsibilities and powers to LGUs, the absence of effective accountability mechanisms allow local chief executives to exercise powers imprudently and thus fail to ensure both economic and human development. Codified rules and legislated frameworks rarely literally translate as intended into the social reality. For this reason, Book IV, Title II, Section 521 of the Local Government Code provides a mandatory review of the code every five years. Two decades and three administrations have passed yet no comprehensive review and amendments have been made on the LGC of 1991. Fortunately, a $250-million loan initiated by the Asian Development Bank on February 13, 2014 will make a review of the 20 year old LGC finally possible. Any review and corresponding amendments to the LGC of 1991 must take into account not only the technical and economic aspect of the code, but especially its political outcomes and consequences. The problems of efficiency in the IRA must be placed in context of a larger patronage structure.

Policy makers must be careful not to focus too much on the economic and technical issues of the LGC; a politically insensitive reform will not yield any desired economic outcome. In the final analysis, while these practical measures may curb unwanted tendencies and patterns at the local level, only a national-level political reform that destroys the systematic patronage structure of the Philippine state can any decentralization framework genuinely achieve its purpose.

References

Arugay, Aries. 2005. "Accountability Deficit in the Philippines: Implications and Prospects for Democratic Consolidation." Philippine Political Science Journal 44 (9) 63-88.

Asian Development Bank. 2005. Country Governance Assesment: Philippines. Manila: Asian Development Bank.

—. 2007. Overview of NGOs and Civil Society: Philippines. Manila: Asian Development Bank.

Atienza, Maria Ela L. 2006. "Local Governments and Devolution in the Philippines." In Philippine Politics and Governance: An Introduction, by Noel Morada and Teresa Encarnacion Tadem, 414-440. Quezon City : University of the Philippines Press.

Bräutigam, Deborah. 2004. The People’s Budget? Politics, Participation and Pro-poor Policy. New York: The United Nations.

Bureau of Local Government Finance. 2001-2009. Statement of Receipts and Expenditures. Quezon City: Bureau of Local Government Finance.

Capuno, Joseph. 2005. The quality of local governance and development under decentralization in the Philippines . Quezon City: University of the Philippines School of Economics.

Diokno, Benjamin E. 2012. "Fiscal decentralization after 20 years: What have we learned? Where do we go from here?" In The Philippine Review of Economics, by Ramon L. Clarete, 9-26. Quezon City: University of the Philippines School of Economics.

Hutchroft, Paul D. 2012. "Re-slicing the pie of patronage: the politics of the internal revenue allotment in the Philippines, 1991-2010." In The Philippine Review of Economics, by Ramon L. Clarete, 109-135. Quezon City: University of the Philippines School of Economics.

Hutchroft, Paul. 2010. "Dreams of redemption: Localist strategies of Reform in the Philippines." In Politics of Change, by Nathan Gilbert Quimpo Yuko Kasuya, 418-454. Pasig: Anvil Publishing, Inc.

Keefer, Philip, and Stuti Khemani. 2003. Democracy, Public Expenditures and the Poor. Washington: The World Bank.

Krafchik, Warren. 2005. Can civil society add value to budget decision-making? A description of civil society budget work. New York: International Budget Project.

Llanto, Gilbert M. 2012. "The assignment of functions and intergovernmental fiscal relations in the Philippines: 20 years after decentralization." In The Philippine Review of Economics, by Ramon L. Clarete, 37-81. Quezon City: University of the Philippines School of Economics.

Monsod, Toby Melissa C., and Emmanuel S. de Dios. 2013. "Scrap pork, empower provinces." Philippine Daily Inquirer, 8 September.

National Statistics Coordination Board. 2012. Annual Per Capita Poverty Threshold, Poverty Incidence and Magnitude of Poor Families, by Region and Province: 1991, 2006, 2009 and 2012. National Statistics Coordination Board.

Snyder, James, and Irene Yackovlev. 2002. Political and Economic Determinants of Government Spending on Social Protection Programs. Cambridge: Massachusets Institute of Technology.

The 8th Congress of the Philippines. 1991. The Local Government Code. Metro Manila: The Republic of the Philippines.

The Lewin Group. 2004. Spending on Social Welfare in Rich and Poor States. Washington: United States Department of Health.

Tidemand, Per. 2010. Political Economy and Governance Analyses of Decentralisation. Denmark: Ministry of Foreign Affairs.

Yilmaz, Serdar, Yakup Beris, and Rodrigo Serrano-Berthet. 2008. Local Government Discretion and Accountability: A Diagnostic Framework for Local Governance. Washington: The World Bank.