Merger myths: a functional analysis of scale economies in New South Wales local government.
In common with a preponderance of other advanced countries, including both unitary states and federations, Australian local government policymakers have employed municipal mergers as an instrument of structural reform (Dollery, Grant and Kortt, 2012). For instance, during the 1990s the South Australian, Tasmanian and Victorian local government systems all experienced council consolidation. Similarly, mergers occurred in New South Wales (NSW) in 2004, Queensland instituted amalgamation in 2007, the Northern Territory underwent wholesale consolidation in 2008, and more recently Western Australia abortively attempted to forcibly merge councils in Greater Perth. Moreover, a further radical episode of forced mergers is currently underway in NSW under the auspices of the Fit for the Future (FFTF) reform process initiated in late 2011.
Advocates of amalgamation frequently argue that it represents an efficacious means of enhancing the operational efficiency of local authorities, improving their administrative and technical capacity, generating cost savings, strengthening financial sustainability and increasing strategic decision-making (Dollery and Robotti, 2008; Faulk and Hicks, 2011). In particular, it is widely asserted that larger local government entities will enjoy significant scale economies attendant upon their size in terms of population (Faulk and Grassmueck, 2012). These claims have been stridently contested in both the conceptual and empirical literature, including the argument that substantial pecuniary savings can be secured through scale economies flowing from larger municipal bodies (Sancton, 2000).
The NSW compulsory council consolidation program, conducted under the FFTF local government reform program, is based inter alia on the premise that the larger local government entities created through forced mergers will exhibit superior performance, due in part to scale economies. However, no empirical evidence has been produced in support of the existence of substantial scale economies in NSW local government. Accordingly, in this paper we test empirically the proposition that significant economies of scale can be reaped in NSW local government by means by municipal mergers using disaggregated 2014 data decomposed by council function.
The analysis of scale economies on the basis of municipal function is based upon the premise that since municipalities conduct multiple production activities, the relationship between expenditure and size is likely to vary based on each activity. Put differently, economies of scale, if they exist, may be 'service-specific' in that councils do not necessarily exhibit scale in all of their activities. This argument is important in light of the advocacy of amalgamation, since amalgamation implies the existence of unanimous and unambiguous scale economies. Accordingly, we investigate this question empirically using multiple regression, within the context of the current debate over municipal mergers in NSW. Our results are stratified by council type and disaggregated these into each of the 11 municipal functional categories in NSW local government. Finally, we also address the empirical miss-specification of proxy variables for output employed in a majority of empirical studies on scale economies in the public sector.
A substantial international empirical literature exists on economies of scale at the local level, which includes a number of empirical papers on scale in NSW local government (see, for example, Dollery et al, 2006; Dollery et al, 2012 for recent surveys of this literature). Various problems have been identified with extant empirical work on scale in NSW municipalities, not least the aggregation of all municipal services and the use of population rather than the number of property tax assessments. The present paper seeks to address some of these problems and thereby offer a more accurate method of estimating of scale by municipal function for defined categories of NSW local authorities.
The paper is divided into five main parts. Section 2 provides a brief discussion of the institutional background to NSW local government and the Fit for the Future structural reform process. Section 3 offers a synoptic account of the empirical measurement of scale economies by scholars in local government. Section 4 presents the empirical strategy followed in the paper and Section 5 outlines the results of the estimation process. The paper ends with some brief concluding remarks in Section 6.
2. STRUCTURAL REFORM IN NEW SOUTH WALES LOCAL GOVERNMENT
The largest of the seven Australian state and territory local government systems, NSW local government is comprised of 152 'general purpose councils', 12 'special purpose councils' and the NSW Aboriginal Land Council. The regulatory structure governing NSW local authorities consists of the NSW Office of Local Government (OLG), the NSW Local Government Grants Commission and supplementary bodies, as well as a Local Government Act (1993). In common with other Australian municipal systems, NSW local government is a 'creature of statute' since the NSW Government possesses almost unlimited regulatory powers over local government.
Relative to most other developed countries, Australian local government fulfils a comparatively limited range of functions, focused largely on 'services to property'. It does not provide many 'services to people', such as education, fire protection and police, which are state government responsibilities. NSW councils provide local infrastructure, including local roads, and local services, such as sewage and solid waste disposal, with some regional councils running local water utilities. NSW local government is funded through property taxes, fees and charges for services, intergovernmental grants, developer charges and various other minor sources. NSW local authorities fall under elected councilors and usually an indirectly elected mayor, who face periodic elections. The operation of councils is conducted by a professional bureaucracy led by a General Manager.
The forced merger program embodied in the Fit for the Future program is derived from recommendations made by an Independent Local Government Review Panel (ILGRP) appointed by the NSW Government in 2011. In its Final Report Revitalizing Local Government--presented in October 2013--the Panel advocated a radical forced merger program, largely centered in Greater Sydney. It was attacked on several counts, not only because of the absence of any empirical basis for its merger recommendations, but also the poor quality of its commissioned research, particularly Assessing Processes and Outcomes of the 2004 Local Government Boundary Changes in NSW by Jeff Tate Consulting (2013).
As we have seen, structural reform through municipal mergers has been attempted in all Australian local government systems. It is generally presented in terms of a 'desperate predicament' polemic (see, for example, Hirschman, 1991) and Fit for the Future is no exception. For instance, in his Parliamentary address outlining FFTF, Minister for Local Government Paul Toole (Hansard, 2014b, p. 399) observed that '[W]e cannot continue as we have done in the past' since 'councils are losing $1 million a day and it is not the way of the future'.
Despite claims by the NSW Government that it was adhering to the recommendations of the Panel, FFTF introduced several new elements into the reform process. For example, the NSW Government subsequently instructed the Independent Pricing and Regulatory Tribunal (IPART) to assess NSW councils' 'fitness for the future'. Under the IPART methodology councils were required to demonstrate that they possessed 'adequate scale and capacity' in addition to ongoing financial sustainability to avoid forced amalgamation. 'Adequate scale and capacity' was not defined but simply characterized on the basis of a set of qualitative criteria largely related to the quality of governance. Furthermore, financial sustainability was subordinated to the amorphously defined 'scale and capacity' criterion such that councils were deemed 'unfit' irrespective of financial sustainability. This has drawn sharp criticism in the scholarly literature (Drew and Dollery, 2015a; Abelson and Joyeux, 2015).
On the 18th December 2015 the NSW Government announced its proposed mergers: Sydney metropolitan councils would fall from 43 to 25, with non-metropolitan councils declining from 109 to 87 entities. These recommended mergers were to be examined by 'Delegates' appointed by the Office of Local Government as part of a public consultation process commenced in January 2016. Delegates were required to make recommendations to the NSW Boundaries Commission which, in turn, would make its own recommendations to the Minister for Local Government.
Almost all of the mergers subsequently accepted by the Minister have been contested by the affected local authorities. In addition, many aspects of the merger process have been disputed and are presently mired in high court proceedings. In many cases, these judicial proceedings have not yet been resolved and it is thus not feasible at present to determine how many forced amalgamations will eventuate.
The FFTF process has generated more problems than it has resolved. For instance, the announcement of an additional $15-20m per amalgamation to merged councils will do little to address the infrastructure maintenance backlog which formed a key aim of FFTF in the first place. Moreover, the uniform allocation of monies across merged councils also will do little to improve municipal fiscal discipline. In addition, over time mergers will imply property tax (i.e. rates) 'harmonization' between amalgamated councils in the newly-merged entities thereby initiating net welfare transfers. Likewise, the NSW Government has announced that the current 'rates trajectory' will remain frozen for four years while IPART reviews the NSW rating system. Over this period, residents within merged entities will pay differential rates for equivalent service provision.
The assessment process for FFTF has been attacked by the local government sector and scholars alike. For example, Revitalizing Local Government has been found to contain substantive errors, to employ unreliable data and to neglect important factors (Drew and Dollery, 2015). In addition, the independence of IPART has been questioned, including evidence of the NSW Premier consulting with IPART prior to the formation of the FFTF policy (NSW Legislative Council Committee Report, 2015). Finally, the arbitrary and changing nature of the assessment criteria has called into question the rigor of the FFTF process.
IPART's findings were based, in part, on commissioned empirical work by EY. In its report, EY conceded that its 'focus is not on the absolute merits of any merger per se' (IPART, 2015, p. 148), but rather only on relative financial merit, in contradiction to the priority of scale and capacity over financial criteria. Further, it conceded the 'limited commercial evidence on the practical impacts of council mergers' (IPART, 2015, p. 148) and made an arbitrary series of assumptions of net present value efficiency savings derived from 'case studies of amalgamation experience elsewhere (mainly New Zealand)' (IPART, 2015, p. 152) which were not identified.
It is more than a little surprising that assumptions underpinning NSW mergers were made using evidence garnered from municipal amalgamations abroad, including in countries where local government has a different relationship with higher tiers of government and operates in a different industrial relations environment. This is even more curious when one considers that IPART decided to neglect the 2007/08 Queensland amalgamation experience despite the strong parallels between the NSW and Queensland local government systems. This may perhaps be ascribed to the fact that the Queensland mergers resulted in higher costs and led to 15 de-amalgamation bids and four actual de-mergers (Drew et al, 2016). In particular, the Queensland case demonstrated that the transactions costs associated with municipal mergers had been greatly underestimated, a problem acknowledged by EY (2015), but not taken into consideration by IPART (2015). Finally, EY (2015, p. 5) conceded that (a) 'econometric analysis does not provide strong support' for municipal mergers and (b) that 'the available empirical evidence on the extent to which local council amalgamations will yield net savings in costs is mixed and tends to vary by activity'. However, cost differentials between different municipal functions were not addressed by EY. This paper addresses this critical gap in the empirical analysis of Fit for the Future by undertaking a functional municipal expenditure analysis for NSW using 2014 data.
3. EMPIRICAL MEASUREMENT OF SCALE ECONOMIES IN LOCAL GOVERNMENT
At its most general, economies of scale refers to the cost advantages which entities derive from size, output or scale of operation, with unit costs of output generally decreasing with increasing scale as fixed costs are spread over greater output. More specifically, scale economies reflect a fall in long-run average and marginal costs arising from an increase in size of an operating unit, such as a local authority (Ferguson, 1969). Economies of scale can be internal to a municipality (i.e. a cost reduction due to management factors) or external (cost reduction due to technological change across a local government system). In this paper, we are concerned with internal scale economies.
In many instances, low levels of production are characterized by diseconomies of scale with rising average costs, followed by scale economies after a given threshold (Dollery et al, 2012). In the local government milieu, scale economies may result from 'higher utilization of fixed assets owned by the council, greater opportunity to exploit the benefits of specialization and discounted bulk-purchasing of inputs' (Dollery et al, 2006, p. 160). However, for some municipal functions, there is no association between municipal size and expenditure, implying that the council expenditure by function may instead be determined by other factors.
Across all types and sizes of municipalities, empirical evidence on economies of scale has been mixed (Drew and Dollery, 2014). For example, Aulich et al (2014) found little evidence of consistent economies of scale in Australian local government. Diseconomies of scale have been identified in Perth (Drew and Dollery, 2014b), Tasmania (Drew et al, 2013) and NSW (Drew et al, 2014; Drew et al, 2015). In addition, attempts to estimate an optimal municipal size have proved elusive, though it has been proposed that many municipalities scheduled for amalgamation may 'already exceed optimal scale' and will 'exhibit decreasing returns to scale' (Drew et al, 2015, p. 1).
These Australian findings have been echoed in empirical work in other countries. For example, in an American context, no significant financial gains from council consolidation were identified by Holcombe and Williams (2009), Couch et al (2004) and Faulk et al (2013). In Canadian local government, McDavid (2002) found that amalgamation in Nova Scotia had resulted in higher costs as well as poorer quality service provision. In Finland, Moisio and Uusitalo (2013) found evidence to suggest that per capita expenditures increased more in merged municipalities than in a comparison group over an 11year period. Dollery, Grant and Kortt (2012, Table 5.1) provide a useful synopsis of the international empirical evidence on sale economies.
In the empirical analysis of scale in local government, it is important to use an appropriate functional unit. Most researchers in the field have employed population as a proxy for size and output, mostly for its ease of use, measurement and interpretation. In addition, population dispersion (Bel and Warner, 2015) and population density (Drew and Dollery, 2014, Holcombe and Wiliams, 2009, Pevcin, 2014) have been deployed.
However, the nature of Australian local government services, focused primarily on 'services to property', means that population size may not be a suitable proxy for local government output, since council services are provided instead to properties (i.e. households, businesses and farmland). In addition, certain demographics are relevant in controlling for segments of the community which exert differing demands for council services (Andrews and Boyne, 2009). This includes age-based considerations (Pevcin, 2014, Drew et al, 2013, Holcombe and Wiliams, 2009), the unemployment rate (Pevcin, 2014, Holcombe and Wiliams, 2009), measures of disadvantage (Holcombe and Wiliams, 2009), income levels of residents (Abelson and Joyeux, 2015, Pevcin, 2014, Holcombe and Wiliams, 2009) and road length (Drew and Dollery, 2014).
Whilst population has dominated the literature as the proxy measure for municipal size, recognition of its limitation has led researchers to seek more appropriate proxy variables. In Australian local government, Drew and Dollery (2014) have demonstrated that the number of households and employing business is a better proxy for output than population per se. However, data on the total of all rate-paying units in a municipality is now available in NSW, denominated as the 'number of assessments'.
Drew and Dollery (2014) have argued that population as a proxy is likely to underestimate the unit cost of production and thus overestimate the point at which economies of scale are exhausted. In addition, number of assessments proxy is especially pertinent in areas that have relatively low population per household, but relatively high concentration of business and/or farming properties. However, one potential limitation of using the number of assessments in NSW is that a small number of entities would not be captured, such as rate exempt properties, including churches, charities, properties owned by higher tiers of government, schools, as well as specific land cultivation discharged from paying rates (Deloitte Access Economics, 2013). Moreover, vacant land is assessable but exerts little pressure on municipal cost functions. An additional complication in appropriately measuring variables for economies of scale relates to possible measurement error on service costs. This includes the allocation of overhead and administrative expenses to particular services, as well as some latitude in how accountants actually allocate costs to different functions (Byrnes and Dollery, 2002), though the relatively large sample size employed is likely to ameliorate such minor measurement errors. Nonetheless, in NSW the number of assessments offers the most appropriate proxy to date for municipal output, given the limited number of rate exempt and vacant properties in most local authorities.
Empirical studies into scale in local government have typically considered the aggregated expenditure of councils. This neglects the fact that councils invariably provide numerous functions, which implies that production functions must be disaggregated based into specific activities, rather than simply aggregate all production (see, for instance, Drew and Dollery, 2014; Pevcin 2014). Councils in NSW operate a range of functions including governance, administration, public order and safety, health, environment, community services and education, housing and community amenities, recreation and culture, mining, manufacturing and construction, transport and communication, and economic affairs. In addition, in NSW there exists a significant difference between 'general-purpose' councils, primarily located outside metropolita n areas, which are responsible for the provision of water and sewerage services, and other councils, which secure these services from regional utilities, like Sydney Water and Hunter Water. It follows that cost functions for each functional category are unlikely to be the same.
Differential scale economies--and hence differential efficiency points, cost functions and cost drivers--may arise because, for instance, the capture of economies of scale in some outputs may be negated by diseconomies in others, since no single size of local government will be able to produce all services at minimum cost (Dollery et al, 2006, Drew and Dollery, 2014b). Accordingly, amalgamation is unlikely to incorporate the least-cost 'size' of municipality for all given municipal functions. This may result in economies of scale in some services and diseconomies of scale in others. Moreover, amalgamation will only be beneficial if scale-potential activity makes up a suitably large mix of the municipal output and only in cases where cost savings from increased scale offset any potential cost involved in effecting a merger and any increases in costs arising from other municipal services.
In the Australian empirical literature, Drew and Dollery (2014) found that only two of ten council functions for Perth councils demonstrated evidence of economies of scale: 'recreation and cultural' and 'governance' expenditures. Within the context of NSW local government, the implication that scale economies apply ubiquitously across all services has been rejected empirically (Dollery et al, 2007) and conceptually (Dollery and Fleming, 2006). Dollery and Fleming (2006) observed that the potential for scale economies on production costs was in decline due to the shift in the nature of production activities of local government: namely, from relatively capital-intensive 'services to property' to comparatively labor-intensive 'services to people'. Moreover, it has been asserted that particular scale characteristics are likely to pertain to specific services (Dollery et al, 2008). In turn, Drew et al (2014) observed that scale economies could only be found in environmental and health services functions in NSW metropolitan councils, and that there is limited overall evidence of scale economies. In addition, studies have considered the potential for scale economies in specific production activities, including waste collection services (Webber et al, 2002), water services (Woodbury and Dollery, 2004) and economic services (Soul, 2000). In the case of waste services, the authors could not offer unequivocal support for the existence of scale economies --and importantly--indicated minimal benefits arising for metropolitan councils in particular. In municipal water service provision, the authors found that half of all councils were producing in the domain of scale diseconomies. Finally, in economic services, net scale diseconomies dominated for municipalities with populations between 100,000 and 316,000 residents.
Within the NSW setting, a variety of empirical methods have been employed, principally multiple regression analysis (see, for instance, Drew et al, 2014) and data envelopment analysis (Drew et al, 2015; Drew et al, 2015b). As we have noted, scale economies must be empirically disentangled from size, density and scope economies (Dollery and Fleming, 2006; Dollery et al, 2012). Irrespective of the empirical methodology employed, claims of substantial scale economies in NSW have largely been refuted by a host of academic studies (see Table 1). This paper adds to the empirical literature on NSW scale economies by examining the multi-product nature of municipalities. In addition, we address the empirical shortcomings in the form of the proxy for output, as well as the appropriate strata of 'general purpose' and 'non-general purpose' councils.
4. EMPIRICAL ANALYSIS OF SCALE IN NSW LOCAL GOVERNMENT
The principal aim of this paper is to investigate whether there is empirical evidence of economics of scale in any, all, or some municipal functions in NSW local government. Economies of scale will be determined by evidence of a statistically significant association between the size of council (in terms of number of assessments) and the expenditure per rate-paying unit. Scale economies refer to a statistically significant association for the downward-sloped portion of a U-shaped cost function between measures of size and functional expenditure per assessment where average costs are falling. For those functional categories where there is a U-shaped functional form, it is proposed that the inflection point represents the optimal size of council jurisdiction for the particular function. In other words, the local minima--if they exist--represent the optimal size of local government for the specific municipal function in question.
We have employed the number of assessments as a measure of council size in preference to population--the standard measure used in the empirical literature--despite recognizing that it is not unproblematic. Indeed, as a proxy for output, the number of assessments is often not the most accurate yardstick. For example, if we are investigating municipal expenditure on recreational areas, then other variables, such as the number of pupils in local schools or the spatial size of recreational areas, are obviously better proxies than simply the number of assessments. Much the same would apply to many other functional categories of municipal activity.
However, our analysis is conducted within the context of contemporary NSW local government which is undergoing structural reform by means of contentious forced mergers. Reform initiatives like the current NSW amalgamation program do not contemplate merging individual local services or specific municipal functions but rather spatial constellations of entire councils. Moreover, enabling legislation is versed in terms of local authorities' per se and not local services. Thus a given measure of size must be employed because real-world merger decisions are invariably made on the basis of a single metric for size applicable to all local authorities. Furthermore, effective political rhetoric in contested arenas, such as forced mergers, almost always utilizes easily communicable simple constructs, such as population size. Indeed, in its current program the NSW Government itself used population as a proxy for size. In this regard, Drew and Dollery (2014) have demonstrated that--in the Australian local government context--the number of assessments is a superior proxy compared to population size, when a single measure is employed.
Municipal expenditure data and demographic data were painstakingly compiled for each NSW local government area. Municipal expenditure data was sourced from Special Schedule 1: Net Cost of Services of the financial statements of each council for 2014. Additional descriptive data was collected from Australian Bureau of Statistics (2014) National Regional Profile report. The data pertaining to the number of residential, farming and business assessments was collected from the NSW Office of Local Government (2014) Time Series Data. The resulting database was analyzed using cross-sectional regression.
To estimate the relationship between local government expenditure and the number of assessments, an ordinary least squares regression model was employed as follows:
E = [alpha] + [[beta].sub.1]S + [[beta].sub.2]X + [mu]
In this equation, E is the log of expenditure per number of assessments, S is a vector of the size variable (number of assessments and the squared term for number of assessments), X is a vector of control variables (percentage of population under 15 years of age, percentage of population over the age of 65, percentage of population identifying as Aboriginal and Torres Strait Islander, percentage of population from a Non-English speaking background, population density, average income, total length of roads) and [mu] represents an independent identically distributed random error term. The principal point of interest was for statistical significance of the S vector. The optimal size of jurisdiction is identified as the number of assessments at which the average cost function is minimized.
5. EMPIRICAL FINDINGS
Economies of scale are first investigated on the basis of aggregate expenditure, with scale economies then examined for each functional activity. All results separately examine 'non-general purpose' and 'general purpose' councils, given their probable disparate production functions.
5.1 EVIDENCE OF OVERALL ECONOMIES OF SCALE
The results in Table 3 suggest that total expenditure per rate-paying unit generally exhibits some evidence of economies of scale, with the average cost function for all NSW councils minimized at a level of approximately 65,000 rate-paying units. Put differently, initial state-wide analysis suggests that most councils currently operate below the average cost curve minima and thus have potential for scale economies. However, when we stratify the regression, into general purpose and non-general purpose councils, in order to untangle the conflation between density and size we find that a statistically significant association only persists for general purpose councils. This finding is consistent with Drew and Dollery (2015) and Drew et al (2014).
5.2 FUNCTIONAL LEVEL EVIDENCE OF ECONOMIES OF SCALE
Table 4 shows that economies of scale exist for municipal production activities derived from Schedule 1 of the audited financial statements. For non-general purpose councils, potential economies of scale only occur in the production of public order services and governance services. These functional categories represent only 6.7% of the municipal expenditure of non-general purpose councils. However, the optimal size for the respective functions varies enormously. Governance expenditure is optimized at 142,027 assessments whereas the production of public order expenditure is optimized at just 5,482 assessments. Accordingly, any attempts to optimize production (and achieve resultant economies of scale) in governance expenditure may incur very high diseconomies of scale for public order. Choosing to merge councils in order to obtain the benefits in governance expenditure across the domain up to 142,027 assessments would imply the reduction of non-general purpose councils from 64 to 21 municipalities. However, creating municipalities of this scale would clearly exacerbate extant diseconomies in the production of public order services. Since non-general purpose councils spend nearly 50% more on average on public order production (4% of total expenditure) than on governance production (2.7% of total expenditure), increasing size beyond 5,482 assessments would result in net increases in per unit expenditure.
For general purpose councils our analysis suggests that only the following functional categories show potential for economies of scale: public order, transport and economic affairs. Similar to the case for non-general purpose councils, the optimal efficient production of these services occurs at dramatically different levels for each of these functions. For instance, the optimal size for general purpose councils in economic affairs production is 36,835 and in public order production it is 244,303. In order to have an optimal size for the production of public order services would necessitate operating with considerable diseconomies of scale in economic affairs production. Since general purpose councils spend on average 74% more of their expenditure on economic affairs (5.4% of total expenditure) than on public order services (3.1% of total expenditure), the scale benefits in public order expenditure are probably less in magnitude than are the diseconomies of scale arising from economic affairs production.
Once regressions are reproduced based on individual functional municipal expenditure categories, it is evident that the size of jurisdiction bears little relationship to expenditure for most municipal services (see Table 4). For some municipal activities, however, there is evidence of economies of scale. However, the optimal scale for each activity occurs at vastly different sizes. In addition to finding that some, but not all, expenditure categories exhibit economies of scale, our results show that: (a) different functional categories may prescribe conflicting optimal size jurisdictions, and (b) the relationship between size and expenditure for any particular functional category may differ for general purpose and non-general purpose councils. In other words, there are no uniform associations between these council strata.
6. CONCLUDING REMARKS
The aim of this study has been to provide a functional municipal expenditure analysis for NSW local government using 2014 data. Given that municipal mergers are expensive, divisive and disruptive, in order for amalgamation to generate net financial gains, efficiency savings must exceed costs. Our findings cast doubt on the potential for net cost savings from the NSW Government's intended mergers. Indeed, for non-general purpose councils, mergers would result in net diseconomies of scale rather than scale economies. A policy implication arising from this paper is that inefficiency is minimized where councils are constituted at the smallest optimal size of any function exhibiting a decreasing average cost curve. To avoid introducing diseconomies of scale, municipalities could then engage in sharing services in other cost functions where increasing size can bring scale benefits.
Amalgamation presents the optimal policy approach where municipalities experience continuously declining average cost curve(s). This study has found, however, that the average cost curve: (a) for most municipal expenditure categories is not driven by size; (b) for those functions where size is the primary driver, there is evidence of a turning point at which economies are exhausted; and (c) in these cases, there are vastly different turning points. As a result, amalgamation is unlikely to realize the significant economic benefits that its proponents contend. Moreover, because there was only evidence of potential economies of scale in a few--and not all--functions, our findings suggest that shared services represent a superior approach to capturing pecuniary savings. This arises because the flexible nature of shared services, relative to boundary mergers, allows municipalities to 'opt-in' to collaboration with other entities to achieve potential scale benefits where they exist, and to 'opt-out' in functions where increasing average costs are present. Accordingly, shared services allows for councils to cooperate and achieve the cost savings in the particular municipal functions where scale exists, while retaining their decision-making autonomy, maintain their local democratic voice, and avoid the disruptive and expensive amalgamation process, while still extracting scale benefits. Moreover, in the case of back-office functions, the use of shared services allows councils to select from a wider range of partner entities than merely their geographical neighbors (Jones and Walker, 2007).
In addition, shared service, resource sharing and other collaborative methods of providing (but not producing) services avoids implicit and explicit transaction costs associated with mergers. These include (a) the moral hazard of debt accumulation ahead of mergers; (b) the allocative inefficiency in resource allocation where there is a 'harmonization of service quality' between merger partners; (c) lesser transparency in large councils; and (d) the exacerbation of principal-agent problems arising from increasing physical and social distance between residents and councils and the associated loss of democratic 'voice'. Moreover, the implementation of mergers typically implies a suspension of democratic representation for extended periods. For instance, the NSW Government has proposed that councils be placed under administration from the announcement of new council boundaries in May 2016 until the postponed council elections in March 2017.
Our results demonstrate a dilemma for policymakers seeking to optimize economic efficiency: how to optimize the cost efficiencies in one functional category while minimizing the inefficiencies which arise in others. This study has found that in order to fully exhaust the economies of scale in a particular municipal function necessarily requires either (a) incurring diseconomies of scale in an alternate municipal function or (b) failing to fully exhaust the scale benefits in at least one municipal function. (a) implies cost inefficiency and deadweight loss in municipal production arising from producing in the domain of increasing average cost, whereas (b) implies that there exists excess capacity along the domain of increasingly efficient production (or declining average cost). Since municipalities should be structured so as to maximize scale benefits, and minimize cost, to maximize the 'return' to rate-paying residents, the exclusivity of the (a) and (b) explains the fundamental paradox that arises from municipal mergers as a policy instrument: boundary reform--by definition --will result in no net cost savings for most municipal functions, and where scale benefits exist in more than one municipal function, the effect of differential cost functions results in either deadweight losses or excess capacity in municipal production.
While our paper has considered the potential efficiency costs and benefits of alternate policy approaches, further empirical investigation should consider producing generalized monetary estimates of these costs and benefits across NSW. In order to advance this empirically, researchers can estimate the marginal cost of production at various sizes of jurisdiction, which may be estimated either service-specifically or generally. This is particularly relevant given that 'opportunity cost' claimed by the NSW Government as applying to councils which have--as yet--refused to merge, may be empirically measured by the marginal cost of production. Moreover, this methodology could be extended to the production of municipal services for which the primary cost driver is a variable other than size. For instance, transport expenditure may be better approximated by road length. In addition, other empirical techniques, such as data envelopment analysis and stochastic frontier analysis, may be profitably employed to provide point estimates of relative technical efficiency of councils. Finally, our study has employed cross-sectional datasets. However, further research could consider conducting a fixed effects panel regression analysis of municipal functional expenditure in order to investigate for generalized findings of dynamic efficiency.
Centre for Local Government
University of New England
Armidale, NSW Australia
Institute of Public Policy and Governance
University of Technology Sydney
Ultimo, NSW Australia
Professor of Economics
University of New England
Armidale, NSW Australia
Abelson, P. and R. Joyeux (2015) "New deveopment: Smoke and mirrors-fallacies in the New South Wales government's views on local government financial capacity", Public Money & Management, 35(4):315-320.
Andrews, R. and G. Boyne (2009) "Size, Structure and Administrative Overheads: An Empirical Analysis of English Local Authorities", Urban Studies, 46(4):739-759.
Aulich, C., G. Sansom and P. McKinlay (2014) "A Fresh Look at Municipal Consolidation in Australia", Local Government Studies, 40(1): 1-20.
Australian Bureau of Statistics, Australian Taxation Revenue 2013-14, released 13/5/15, available at www.abs.gov.au/ausstats/abs@.nsf/mf/5506.0.
Baird, M. (2015) Alan Jones Breakfast Show 2GB [Interview] (Friday 23rd October 2015).
Bel, G. and M. Warner (2015) "Inter-Municipal Cooperation and Costs: Expectations and Evidence", Public Administration, 93(1):52-67.
Byrnes, J. and B. Dollery (2002) "Do Economies of Scale Exist in Australian Local Government? A Review of the Research Evidence", Urban Policy and Research, 20(4):391-414.
Couch, J., C. Gossett, B. King and J. Parris (2004) "Economies of scale and the provision of public goods by municipalities", Journal of Economics and Economic Education Research, 5(3):69-79.
Deloitte Access Economics (2013) Review of Local Government Rating Exemption Provisions.
Dollery, B., B. Grant and M. Kortt (2012) Councils in Cooperation: Shared Services and Australian Local Government. Sydney: The Federation Press.
Dollery, B., L. Crase and A. Johnson (2006) Australian Local Government Economics. Sydney: UNSW Press.
Dollery, B., J. Byrnes and P. Allan (2007) "Optimal Structural Reform in Australian Local Government: An Empirical Analysis of Economies of Scale by Council Function in New South Wales", Urban Policy and Research, 25(4): 473-486.
Dollery, B. E., J. Byrnes and L. Crase (2008) "Australian Local Government Amalgamation: A Conceptual Analysis of Population Size and Scale Economies in Municipal Service Provision", Australasian Journal of Regional Studies, 14(2):167-175.
Dollery, B. E. and E. Fleming (2006) "A Conceptual Note on Scale Economies, Size Economies and Scope Economies in Australian Local Government", Urban Policy and Research, 24(2) 271-282.
Dollery, B. E. and L. Robotti (eds.) (2008) The Theory and Practice of Local Government Reform, Edward Elgar: Cheltenham.
Faulk, D., P. Schaal and C.D. Taylor (2013) "How Does Local Government Amalgamation Affect Spending? Evidence from Louisville, Kentucky", Public Finanace & Management, 13(2): 80-98.
Faulk, D. and M. Hicks (2011) Local Government Consolidation in the United States, Cambria Press: Amherst.
Drew, J., B. Grant and N. Campbell (2016) "Progressive and Reactionary Rhetoric in the Municipal Reform Debate in New South Wales Australia", Australian Journal of Political Science, DOI: 10.1080/10361146.2016.1154926.(22/1/16).
Drew, J., M. Kortt and B. Dollery (2013) "A Cautionary Tale: Council Amalgamation in Tasmania and the Deloitte Access Economics Report", Australian Journal of Public Administration, 72(1):55-65.
Drew, J. and B. Dollery (2014) "Keeping it in-house: Households versus Population as Alternative Proxies for Local Government Output", Australian Journal of Public Administration, 73(2):235-246.
Drew, J. and B. Dollery (2014b) "Would Bigger Councils Yield Scale Economies in the Greater Perth Metropolitan Region? A Critique of the Metropolitan Local Government Review for Perth Local Government", Australian Journal of Public Administration, 73(1): 128-137.
Drew, J. and B. Dollery (2015) "Less Haste, More Speed: The Fit for the Future Reform Program in New South Wales Local Government", Australian Journal of Public Administration, 75(1):1-11.
Drew, J., M. Kortt and B. Dollery (2014) "Economies of Scale and Local Government Expenditure: Evidence from Australia", Administration & Society, 46(6):632-653.
Drew, J., M. Kortt and B. Dollery (2015) "No Aladdin's Cave in New South Wales? Local Government Amalgamation, Scale Economies and Data Envelopment Analysis Specification", Administration & Society, DOI: 10.1177/0095399715581045:1-21.
Drew, J., M. Kortt and B. Dollery (2015b) "What Determines Efficiency in Local Government? A DEA Analysis of NSW Local Government", Economic Papers, 34 (4):243-256.
Drew, J. and B. Dollery (2015 a) Submission to the General Purpose Standing Committee No. 6 Inquiry into Local Government in New South Wales.
Ernst and Young (2015) Fit for the Future: Review of Business Case Estimates of Merger Net Benefits for Sydney Metropolitan Councils, October 2015.
Faulk, D and G. Grassmueck (2012) "City-county Consolidation and Local Government Expenditures", State and Local Government Review, 44(3):196-205.
Ferguson, C. E. (1969) Neoclassical Theory of Production and Distribution, Cambridge: Cambridge University Press.
Fogarty, J. and A. Mugera (2013) "Local Government Efficiency: Evidence from Western Australia", Australian Economic Review, 46(3):300-311.
Hirschman, A. (1991) The Rhetoric of Reaction: Perversity, Futility, Jeopardy. Cambridge: Belknap Press of Harvard University Press.
Holcombe, R. and D. Wiliams (2009) "Are there Economies of Scale in Municipal Government Expenditures?", Public Finance and Management, 9(3):416-438.
Independent Local Government Review Panel (ILGRP) (2013) Revitalising Local Government: Final Report of the NSW Independent Local Government Review Panel, ILGRP.
ILGRP (2013 a) Future Directions of NSW Local Government--Twenty Essential Steps, Independent Local Government Review: Sydney.
Independent Pricing and Regulatory Tribunal (IPART) NSW (2015) Assessment of Council Fit for the Future Proposals Final Report, October 2015.
Jeff Tate Consulting (2013) Assessing Processes and Outcomes of 2004 Local Government Boundary Changes in NSW, Jeff Tate Consulting, McLaren Vale.
Jones, S. and R. Walker (2007) "Explanators of Local Government Distress", Abacus, 43 (3):396-418.
McDavid, J. (2002) "The impacts of amalgamation on police services in the Halifax Regional Municipality", Canadian Public Administration, 45(4):538-565.
Moisio, A. and R. Uusitalo (2013) "The Impact of Municipal Mergers on Local Public Expenditures in Finland", Public Finance and Management, 13(3):148-166.
NSW Legislative Council General Standing Committee No. 6. (2015) Senate Standing Committee Media Release: Committee finds local government reforms 'unfit', 29 October 2015
Pevcin, P. (2014) "Local Government Cost Function: Case Study Analysis for Slovenian Municipalities", Zagreb International Review of Economics & Business, 17(1):79-86.
Sancton, A. (2000) Merger mania. Quebec: McGill-Queens University Press.
Soul, S. (2000) "Population Size and Political Performance of Local Government Jurisdictions", Unpublished PhD thesis, Lismore: Southern Cross University.
State Parliament of NSW [Hansard] (2014b) Legislative Assembly 11 September 2014 [full transcript], accessed 8 October 2015, (https://www.parliament.nsw.gov.au/prod/parlment/hanstrans.nsf/V3ByKeyL A20140911/$File/LA20140911.pdf).
Webber, A., B. Dollery and J. Byrnes (2002) "Measuring Economies of Scale in Australian Local Government: the case of domestic waste collection in NSW", Australiasian Journal of Regional Studies, 8(2):201-218.
Woodbury, K. and B. Dollery (2004) "Efficiency Measurement in Australian Local Government: The Case of New South Wales Municipal Water Services", Policy Research, 21 (5):615-636.
Table 1. Empirical Analysis of Scale Economies in Australian Local Government Author Setting Empirical Principal Findings Technique Anlich et al Australia Case studies Little evidence (2014) wide and interviews that with sector amalgamation professionals will produce scale economies Dollery et al NSW Interviews with Rejects that (2007) Council GMs scale economies apply across all functions Drew et al NSW DEA Demonstrated (2015) that 20 of 23 proposed mergers will be over-scaled and exhibit diseconomies of scale Drew et al NSW Multiple Little overall (2014) regression evidence of scale economies Drew et al QLD Multiple 84% of (2016) regression residents in councils operating with diseconomies of scale Drew et al TAS Multiple No evidence of (2013) regression scale economies Drew and Perth Multiple Scale economies Dollery (2014b) regression are function specific and evident in 2 out of 10 functions Drew and NSW Multiple Rejected scale Dollery (2015) regression economies within the Sydney region, though did identify economies of density Fogarty and WA DEA Mixed evidence Mugera (2013) on scale economies Woodbury and NSW DEA Half of all Dollery (2004) councils experiencing diseconomies in water services production Table 2. Variables for Functional Analysis of Economies of Scale, NSW 2014 Variable Definition State Non- General Mean General Purpose Purpose Council Council Total Total of all 4284.52 2795.02 5367.79 expenditure per expenditures from assessment (ln) all functions Governance Includes: costs 92.12 68.17 109.54 expenditure per relating to assessment (ln) council's role as a component of democratic government Administration Includes: costs 567.95 488.67 625.6 expenditure per related to corporate assessment (ln) support and other support services Public order Includes: fire 142.50 113.01 164.94 expenditure per protection, animal assessment (ln) control, beach control, enforcement of local regulation, emergency services Health Includes: 39.11 24.18 49.97 expenditure per administration and assessment (ln) inspections, immunizations, food control, insect control, noxious plants, health services Environmental Includes: 446.53 441.29 450.34 Services environmental expenditure per protection, solid assessment (ln) waste management, street cleaning, drainage, storm water management Community and Includes: family day 223.37 168.66 263.16 education care, child care, services youth services, expenditure per aged and disabled, assessment (ln) migrant services, Aboriginal services, education Housing and Includes: housing, 187.93 174.28 198.86 community town planning, amenities per domestic waste assessment (ln) management services, sanitation and garbage, public cemeteries, public conveniences, other community amenities Recreational Includes: public 499.94 466.05 524.59 and Cultural libraries, museums, expenditure per art galleries, assessment (ln) community centers, public halls, swimming pools, sporting grounds, parks and gardens Mining, Includes: building 61.76 45.28 73.75 manufacturing control, abattoirs, and quarries and pits construction assessment (ln) expenditure per Transportation Includes: urban 1302.20 656.45 1771.84 and roads, rural roads, communication bridges, expenditure per foot-paths, assessment (ln) aerodromes, parking areas, bus shelters, street lighting Economic Includes: camping 216.26 107.22 295.57 affairs areas, caravan expenditure per parks, tourism and assessment (ln) area promotion, industrial development promotion, real estate developments, other business undertakings Number of the sum of total 19883.5 31711.6 11281.2 assessments residential, (ln) business and farming rate-paying units Population population per 786.1 1813.7 30.2 density (ln) squared kilometer Population percentage of 25.6 24.7 26.2 under 15 years population under the (%) age of 15 Population over percentage of 23.7 21.3 25.4 65 years (%) population over the age of 65 ATSI (%) percentage of 5.3 2.2 7.6 population of Aboriginal or Torres Strait Islander background NESB(%) percentage of 11 21.4 3.5 population from non-English speaking background Average wage average taxable 49070 57093.3 43234.8 (ln) Length of income of residents 1089.7 665.3 1398.4 road (km) in dollars total length of road surface in kilometers per local government area Table 3. Overall Economies of Scale, NSW 2014 Council Strata Total Expenditure Entire State Non-General General Purpose Purpose Assessments (ln) -0.611 (+) Assessments (-1.68) -0.635 (-1.28) -1.382 * (-2.44) squared (ln) Coefficient of 0.028 (1.45) 0.027 (1.08) 0.073 * (2.4) Determination 60.89 60.29 47.62 Indicative 65,571 Not Significant 13,001 Optimal Size (+) p < 0.10, * p < 0.05, ** p < 0.01 Table 4. Functional Level Economies of Scale by Council Strata, NSW 2014 Function Governance Administration Council Strata Non- GP Non- GP GP GP Assessments (ln) -3.808 * 1.257 -1.56 1.646 (-1.88) (-1.02) (-0.88) (-0.91) Assessments 0.161 0.032 0.078 0.084 squared (ln) (1.54) (0.49) (0.87) (0.86) Coefficient of 41.92 33.43 3.1 24.14 Determination Indicative Optimal 142,02 Size 7 N/A N/A N/A Function Public Order Health Council Strata Non- GP Non- GP GP GP Assessments (ln) -1.481 -1.334 (+) -3.059 0.562 (-1.51) (-1.84) (-1.02) (-0.23) Assessments 0.086* 0.054 0.089 -0.261 squared (ln) (1.73) (1.39) (0.57) (-0.2) Coefficient of 43.52 39.08 22.52 24.34 Determination Indicative Optimal Size 5,482 244,303 N/A N/A Function Environmental Community and Education Council Strata Non- GP Non- GP GP GP Assessments (ln) 0.608 -1.625 0.908 0.569 (0.81) (-0.25) (-0.35) (0.21) Assessments 0.021 0.016 0.03 -0.004 squared (ln) (-0.56) (0.44) (0.23) (-0.03) Coefficient of 39.1 35.04 6.43 5.53 Determination Indicative Optimal Size N/A N/A N/A N/A Function Housing and Recreation Amenities Council Strata Non- GP Non- GP GP GP Assessments (ln) 0.581 0.109 -0.623 -0.716 (0.5) (0.13) (-0.93) (-1.25) Assessments -0.019 -0.003 0.032 0.321 squared (ln) (-0.31) (-0.07) (0.93) (1.05) Coefficient of 22.13 8.83 30.68 30.42 Determination Indicative Optimal Size N/A N/A N/A N/A Function Mining , Transport manufacturing, construction Council Strata Non- GP Non- GP GP Assessments (ln) -1.561 -0.79 1.044 (-0.65) (-0.31) (0.98) Assessments 0.115 0.034 -0.059 squared (ln) (0.94) (0.81) (-1.1) Coefficient of 40.38 17.67 75.05 Determination Indicative Optimal Size N/A N/A N/A Function Transport Economic Affairs Council Strata GP Non- GP GP Assessments (ln) -1.73** -0.729 -2.138 (+) (-2.81) (-0.24) (-1.78) Assessments 0.077** 0.014 0.102 squared (ln) (2.34) (0.09) (1.58) Coefficient of 77.32 49.92 29.58 Determination Indicative Optimal Size 74,810 N/A 36,835 (+) p < 0.10, * p < 0.05, ** p < 0.01
|Printer friendly Cite/link Email Feedback|
|Author:||Fahey, Glenn; Drew, Joseph; Dollery, Brian|
|Publication:||Public Finance and Management|
|Date:||Sep 22, 2016|
|Previous Article:||Corruption, regime type, and economic growth.|
|Next Article:||The effects of double tax treaties for developing countries. A case study of Austria's double tax treaty network.|