Local Government Reform as State Building: What the Polish Case Says About “Decentralization”

Abstract

Poland has become one of the more decentralized states in Europe. Local governments now control a third of all public expenditures. They have also delivered the goods modernizing the country’s infrastructure and restructuring its schools. This success cannot be attributed to widespread civic engagement because decentralization in Poland was clearly a “revolution from above.” Nor can it be attributed to the implementation of rules typically thought to enhance accountability in decentralized polities because Polish local governments do not finance themselves, and many of their responsibilities remain poorly defined. Instead, Poland’s success is due to an array of institutions designed to train, professionalize, discipline, and empower newly elected local elites. These institutions suggest that creating effective local governments may lie less in “getting the rules right” or in empowering citizens to participate in their own governance, than in institutions that encourage local governments to monitor themselves while embedding them in the regulatory structure of the state.

Introduction

In the 1980 and 1990s, as communism collapsed and authoritarian regimes transitioned toward democracy, “decentralization” rose to the top of the agenda for governance reform. And for a moment, everybody seemed to agree that putting power and money closer to citizens was a “good thing.” But 30 years, and numerous decentralization efforts later, nobody is so sure. Now the more frequent question is why, if decentralization is such a good thing, are there so few places where it has obviously worked (Treisman 2007)?

Here, Poland constitutes an exceptional case. Subnational governments control over a third of all public expenditures and a remarkable 70 % of public investment. They have also delivered the goods, transforming the country’s environmental infrastructure, transport systems, urban spaces, and perhaps most strikingly its schools. Indeed, Polish schools now rank near the top of the league tables for quality and equity as measured by international tests (OECD 2009, 2012).

What explains Poland’s success? And does this success tell us anything useful about “decentralization” and good governance? In the following, I suggest answers to both questions by examining Poland’s success against the background of the dominant discourses used to justify decentralization efforts and the explanations of why these efforts have so often been disappointing.

The first discourse comes from the literature on fiscal federalism. This literature maintains that decentralization should produce better governance if the functions of different levels of government can be clearly separated from each other, if the benefit areas of local public services can be aligned with the jurisdictions that provide them, and if local services are paid for by local taxes. Under these conditions, the argument runs, decentralization will promote good governance by giving citizens the means (elections), the incentives (taxes), and the information (clearly delineated responsibilities) necessary to hold politicians accountable for their performance (Musgrave 1959; Oates 1972; Ter Minassian 1997).

In this discourse, the inability to get these rules right explains decentralization’s disappointments and is generally attributed to a failure of “political will” (Smoke 2001). More particularly, difficulties making local governments pay for themselves through local taxes is said to produce transfer dependency, fiscal irresponsibility, and macro-economic instability while the inability to clearly separate functions across levels of government is thought to confuse the taxpaying electorate (Tanzi 2002; Rodden 2005). Similarly, the misalignment of benefit areas with political jurisdictions is held responsible for negative spillovers, fragmentation, and political gridlock (Prud’homme 1995; Treisman 2007).

The second discourse about decentralization is less concerned with rules than empowerment. If citizens can be equipped with the capacities, resources, and powers necessary to govern themselves, then decentralization should widen the public sphere, deepen democracy, and improve public services (Cheema and Rondinelli 2007; Connerly 2009; Fung and Wright 2003). Here, decentralization’s disappointments are usually attributed to the intractability of existing power relationships: To the ease with which dominant elites capture or subvert local governments (Bardhan and Mookerjee 2006; Gibson 2012) and/or to the difficulties of capacitating the disenfranchised (McGee and Gaventz 2010; Fung and Wright 2003).

The success of decentralization in Poland cannot easily be understood within either of these frameworks. On the one hand, the architects of Poland’s reforms ignored many of fiscal federalism’s central policy prescriptions, making local governments heavily dependent on transfers and leaving the division of responsibilities in many areas—most notably in education—profoundly confused. On the other hand, there is little evidence that the process was either driven forward by or has resulted in widespread civic engagement. Instead, a small group of policy makers designed and implemented the reforms from above. Moreover, they had modest expectations about citizen participation—expectations that have largely proved well-founded.

So how is it that local government reform in Poland has worked so well in the absence of either the intergovernmental rules or the popular engagement that the dominant literatures place at the foundation of both decentralization’s normative virtues and its practical difficulties? Or put another way, why have Polish local governments not descended into a miasma of fiscal imprudence and rent-seeking if neither the rules governing the system nor the citizens occupying it can be reasonably identified as the primary agents of accountability?

In the following, I argue that Poland’s success is the product of a remarkably self-conscious strategy of institution building by a group of policy makers who shared a specific vision of what local government reform was about. Unlike many of their counterparts in the region (and beyond), they did not primarily regard local governments as repositories of democratic virtue or vehicles for direct civic participation. Nor did they think of them as quasi-autonomous fiscal units in which the alignment of electoral jurisdictions, benefit areas, and tax powers would generate a virtuous circle of good governance.

Instead, they saw local governments as functional components of a single, national system of public administration. They wanted to make local governments responsible for the vast majority of day-to-day public services because they were convinced that this was the only way to ensure that the national government would focus on larger questions of strategy, policy, and law. For them, in other words, “decentralization” was the foundation of a broader state-building strategy.

These reformers assumed that there was little popular demand for local governments and that the capacity to run them was weak. They also believed that deconstructing communism would be a long struggle. This, I argue, led them to construct an array of institutions designed to train and professionalize newly elected local officials and to collectively embed them in the state’s regulatory and policy-making processes. Horizontally, these institutions mitigated the dangers of local capture and rent-seeking by fostering new norms and standards and by forcing local officials to police each-others’ behavior, thus at least partially compensating for the lack of an engaged citizenry. Vertically, they blurred the boundaries between levels of government and made possible the continual adjustment of intergovernmental relations—adjustments that allowed decentralization to unfold as a process, and not as a one-sided attempt to instantiate rules thought to ensure subnational accountability and good governance.

The story I tell is thus about an elite group of state builders whose ideas informed the creation of institutions that professionalized, disciplined, and empowered a new class of democratically elected local officials to govern effectively. In making this argument, I am not denying that much of the impetuous for the rapid creation of democratically elected local governments was narrowly political. After all, almost everywhere in the region elections for newly constituted local governments were held within 2 years of communism’s collapse, and almost everywhere national reformers saw them as a powerful tool for defanging old foes. Nonetheless, Poland’s reforms stand out for their depth, resilience, and effectiveness, and their success cannot easily be attributed to a motivation that was widely shared elsewhere.

Equally importantly, I am not denying the importance of Solidarity either as a trade union or as a social movement (Ekiert and Kubik 2001). Indeed, the fact that the union was (unusually) organized along regional and not branch lines (Ost 1991) and that many of its activists turned—as we shall see—to local governments after 1990, undoubtedly had much to do with Poland’s success.

Nonetheless, what is striking about Poland’s success is that the ideas that informed it were not predicated on strong expectations of widespread civic engagement. Indeed, in many ways, the institutions that the architects of the reforms built can best be understood as a mechanism designed to create responsible local elites in the presumed absence of an engaged citizenry and no faith in the tax-based accountability of fiscal federalism. Moreover, there is little strong evidence that this basic presumption was wrong: Polish post-communist civil society seems to be similar to those in the rest of the region inasmuch as the Poles continue to score at the bottom of the post-communist barrel on measures like trust in government, trust in others, membership in associations, and participation in public processes and events (Bernhard 1996; Kramer 2002; Howard 2003; Tworecki 2008). Again, this is not to deny that the history of popular resistance to communism left Poland in an advantageous position in 1990 by providing the architects of its reforms with a pool of committed activists ready to go into the local governments they were creating. But saying this is very different from attributing the success of the process to an unusually engaged citizenry or a particularly “strong civil society” (Mielczarek 2012).

In the following, I proceed in four steps. First, I outline the reform’s most striking achievements in order to establish that something exceptional has taken place in Poland and that little of it has to do with either fiscal federalism’s rules or popular engagement. Second, I examine how the architects of the reforms understood decentralization and how this understanding informed the institutions they created. Third, I illustrate how these institutions worked to foster new professional norms, promote horizontal accountability, and facilitate the continuous adjustment of intergovernmental relations. Finally, I discuss those features of the Polish case that may be useful when considering “decentralization” in other post-authoritarian contexts.

Poland as a Decentralized Polity

Decentralization in Poland was carried out in two phases, the first in 1990 and the second in 1999. In March 1990, the national government passed the Law on Local Government (LLG) and in May held elections for c. 2500 municipal and communal governments called Gmina (Sejm 1990). The LLG made Gmina responsible for all the basal metabolic functions of urban life¹ as well as for preschool and primary education, ambulatory health care, and some welfare services.

Between 1993 and 1997, efforts to create a second tier of county governments faltered. In 1998, however, Solidarity-affiliated parties retook parliament and within a year completed the so-called Second Phase of decentralization (Sejm 1998). In rural areas, 314 county-level governments were created (Powiat ziemski, Rural Counties) while the 66 largest cities were made Cities with County Rights (Miasto na prawach powiatu, Large Cities). Rural Counties and Large Cities were assigned responsibility for secondary education, county-level transport, and the maintenance of hospital facilities.

Forty-nine województwa were also consolidated into 16 democratically elected regional governments. These “Self-Governing Regions” coordinate development planning and have some infrastructure responsibilities, but deliver almost no day-to-day public services. They “cohabit” with 16 deconcentrated units of the national government which run the police, courts, and a number of inspectorates. But they play a limited role in planning and deliver no other public services.

Despite the creation of Rural Counties and Self-Governing Regions, the most important level of subnational government remains municipal. Gmina and Large Cities control two-thirds of all subnational expenditure (Fig. 1) and deliver most local services. They are financially independent of Rural Counties and Self-Governing Regions, and receive their transfers directly from the state budget.² Polish decentralization is thus best understood as deep “municipalization” within a unitary state, with the 66 Large Cities constituting the backbone of the system.

Fig. 1

Local government expenditures by level of government (billion PLN). GUS/BDL http://www.stat.gov.pl/bdl/app/strona.html?p_name=indeks Own calculations

Figure 2 shows subnational expenditures as a share of GDP and of total public expenditures for the countries that joined the EU since 2004, as well as for those European countries in which subnational governments control a higher share of expenditure than their Polish counterparts. Polish subnational governments play a substantially larger role than those in other new EU member-states (and a much greater role than those in South-East Europe) (NALAS 2015). Only in the Netherlands, Europe’s federations, and the unitary states of Scandinavia do subnational governments control a larger share of the public purse. So Poland is not only the most decentralized post-communist country in Europe but also one of Europe’s more decentralized unitary states.

Fig. 2

Public expenditures as a percent of GDP and subnational expenditures as percent of GDP and of public expenditures (2010–2012 average). EuroStat: http://epp.eurostat.ec.europa.eu/portal/page/portal/statistics/search_database *Federations

All Polish subnational governments depend heavily on national government transfers and only Large Cities derive more than 35 % of their revenue from own-sources (Fig. 3). In Gmina and Large Cities, the most important transfers come from shared taxes and the education and equalization components of the General Grant. All subnational governments get a percentage of the personal income tax (PIT) generated in their jurisdictions, and all but Gmina get a percentage of corporate income tax (CIT). But these shares differ for each type. They have also been adjusted upward following the devolution of new functions and/or negotiations with the national government. Currently, 50 % of the national yield of PIT and 22 % of CIT are returned to subnational governments on an origin basis (Sejm 2003). Here it is useful to note that origin-based tax sharing creates a direct budgetary incentive for local governments to promote job growth and to work with the national government to legalize the grey economy.

Fig. 3

Shares of subnational revenue by type (2014). GUS/BDL http://www.stat.gov.pl/bdl/app/strona.html?p_name=indeks Own calculations

Tax sharing also anchors the equalization system: Jurisdictions whose per capita revenues from shared taxes are less than 90 % of the national average receive a grant equal to 85 % of the difference between the per capita yield of these taxes in their jurisdictions and 90 % of the national average. This system has provided even very poor jurisdictions with reasonable revenues. It is also transparent and simple to administer because it does not use politically contentious proxy measures of relative wealth.

Between c. 20 and 35 % of the revenues of Gmina, Rural Counties, and Large Cities come from the education component of the General Grant. It is calculated on the basis of the number of (weighted) pupils attending schools in a jurisdiction. Rural Gmina receive about 30 % more per pupil than urban Gmina because rural schools have smaller classes. As a result, this component of the grant has also improved the horizontal equity of the system because Rural Gmina tend to be poorer than urban ones (Herbst, Herczyński, and Levitas 2009).

Both Gmina and Large Cities derive c. 12 % of their revenues from an area-based property tax. This is the most important local tax and nationally yields about 1.1 % of GDP. This equals the EU average, and is more than double the average for South-East Europe (NALAS 2015).³ But the relatively good performance of the property tax does not mean that local governments are really taxing citizens. Instead, 85 % of the tax comes from businesses (Przekopiak 2009).

Gmina and Large Cities generate about 5 % of their income from the sale or rental of municipal property. Comparative data on local asset revenue is hard to find, but Poland is certainly a high-flyer. This is because reformers insisted that state-owned land be rapidly transferred to local governments, and because the Holocaust and the post-war expulsion of Germans from Silesia meant large amounts of real-estate were available for municipalization.⁴ The sale and rental of these assets put municipalities at the front of so-called “small privatization.” This was important because the privatization of large state-owned enterprises went slowly (Potkański 2013; Levitas 1992). It also provided municipalities with a powerful tool for controlling urban land-use (Panko 1995).

Poland also has the deepest subnational capital market in post-communist Europe, including an unusually vibrant bond market (Levitas and Kopańska 2003). Between 1995 and 2009, subnational government debt tripled and now exceeds 15 billion euros (GUS 2012 p. 49). Over 80 % of subnational governments have borrowed and only a handful has exceeded their legal limits or been forced to adjust their payment schedules. Thus, despite their “transfer dependency,” local governments have exhibited little of the imprudence expected by fiscal federalists and which indeed is found often enough in practice.

Poland’s intergovernmental finance system has provided subnational governments with what the literature calls “adequate, equitable, and predictable revenues.” The best evidence for this is that between 1991 and 2009, 20 to 25 % of subnational expenditure went to investment (GUS, own calculations). This rate is particularly remarkable because it is from an elevated base given the high operating costs of schools (c. 30 % of expenditure). Equally important, investment rates do not differ significantly between urban and rural jurisdictions, underscoring the basic equity of the system. Most strikingly, for the last 20 years, Polish subnational governments have spent much more on investment as a percentage of GDP—2.5 %—than their counterparts elsewhere in the EU (Fig. 4), accounting for a higher share of public investment (63 %) than subnational governments in both new (44 %) and old (58 %) member-states. In short, subnational governments have played an outsized role in building the new Polish state.

Fig. 4

Public investment by level of government as percentage of GDP (average 1995–2012). EuroStat: http://epp.eurostat.ec.europa.eu/portal/page/portal/statistics/search_database *Federations; CAPS NewEU10

During the 1990s, local governments built close to 2000 biological sewage-treatment plants and tripled the number of household connections. After 1999, they focused on transport, doubling the length of hardened roads (Wspólnota 2010). They have also consistently devoted 8–14 % of investment spending on education (GUS/BDL, own calculations).

Gmina and Large Cities now manage c. 35,000 schools and non-school educational facilities, consistently spending c. 40 % more on them than they receive from the national government (Herbst, Herczyński and Levitas 2009, pp. 103–106).⁵ They have used this additional spending to dramatically improve facilities, double preschool enrollment from 18 to 36 % nationally (c. 80 % in Large Cities, Swianiewicz 2012), and to support substantial amounts of additional teaching time.

Moreover, they have done this in the face of a 27 % drop in enrollment between 2000 and 2010. Demographic decline has forced local governments to close 20 % of all facilities, restructure school networks and transport systems, and reduce the teaching force by 6 %. The gap between falling enrollment and employment has pushed class sizes to the low end of the OECD-spectrum and generated much handwringing about lost efficiency. But given how difficult it is to fire teachers and close schools, what is much more remarkable is the extent of the restructuring (Herbst and Levitas 2012).

Most importantly, decentralization has been accompanied by dramatic improvements in student scores on international tests like PISA and TIMSS. Indeed, Poland’s gains have been among the highest in the world—particularly for poorer students⁶—and have made the country something of a poster child for education reform (Ripley 2013). And while it is impossible to attribute these gains to any single factor, it seems clear—as the authors of the 2009 PISA report put it—that:

It is hard to argue that the devolution of schooling to local governments has thus far been anything other than a positive development for Polish education (OECD 2009, p. 34).

Local Government Reform as State Building

Framing Ideas

In the winter of 1989, the Polish United Workers Party and the still illegal trade union “Solidarity” sat down at the so-called “Roundtable Talks.” The Talks led to the re-legalization of the union and to partially free parliamentary elections in June 1989. By summer’s end, communism had collapsed and by March 1990, Poland had held elections for 2500 Gmina.

The breakneck speed with which Gmina were created suggests that decentralization was a high priority within the opposition. But at least initially, it was of only marginal interest. To be sure, the opposition had used the term “Self-Governing Republic” (Rzeczpospolita Samorządowa) since the late 1970s. But the phrase was not associated with local governments. Instead, it referred to Workers Councils or was understood as a call for civil society to organize itself outside of the communist state. Consequently, while Solidarity’s 1981 programmatic declaration mentioned local government, its 20th point read: “Authentic Workers Councils will be the foundation of the Self-Governing Republic” (Regulski 2000 pp. 17–38).

Thus, during the preparation for the talks, the union’s main negotiators did not put local governments on the agenda. Jerzy Regulski—the acknowledged “father” of the reforms (Paszyński 1996)—explained the oversight bluntly: “At the time, the question of rebuilding local governments was not considered a significant issue either by the existing political authorities or within the opposition community (środowiska)” (Regulski 2000 p. 48). Indeed, it was only at the last minute that Regulski and his collaborators placed the issue on the agenda and established a subcommittee for them. Michał Kulesza, later the Plenipotentiary for Local Government, was so galled by this indifference that his blurb for the publication “Who’s Who at the Roundtable” read:

It is shocking that the opposition has attached so little importance to the issue of local government, leaving it to the “experts” and not understanding that human affairs are decided equally at the place of work as in the place of residence –which is the real field for the creation of networks (środowiska) and political elites (italics mine) (Kulesza 1993 p. 37). ⁷

But while most of the opposition was uninterested in local government, a small group of geographers and administrative lawyers had been working on the question for years. During the 1980s, the group researched the local operation of the party state, closely observed decentralization reforms in Denmark and Sweden, and wrote widely about the need to restore Gmina in Poland. Indeed, by the time communism collapsed, the group can reasonably be called an “epistemological community” that shared a foundational belief that whatever Solidarity’s future might be, the locus of politics in Poland had to be shifted from factory to home.⁸

This community insisted that not only did local governments have to be democratically elected but they also had to have legal identities, independent budgets, property rights, and control over their personnel. They also clearly saw local governments as a tool for dismantling the communist state, a position that the new government quickly adopted. Nonetheless, they protested when colleagues characterized local governments “only as a form of organized cooperation between people and not as elements of public administration” (Regulski 2000 p. 38). And while they believed that bringing power closer to the people would lead to better public services, they did not see the primary virtues of local government in terms of citizen participation. Instead, they identified newly elected local elites as the primary agents of transformation. Kulesza wrote:

In Poland today the issue is not about widespread, everyday participation in local affairs. People are really exhausted. Besides, there are few places in the civilized world in which permanent social and political engagement is expected from citizens. No, that is what elections and democratic representatives are for….it is they –local politicians—who must be exceptionally active and take the results of their work to the electorate for legitimation (Kulesza 2008 p. 28).

This sober assessment ran against regional trends where a “Tocquevillian myth of localism flourished in which local self-government was to be the incarnation of civil society, and everything the regime was not” (Campbell and Coulson 2006). It also kept Polish reformers from allowing every village to declare itself a Gmina. As a result, Poland avoided the jurisdictional fragmentation that has haunted decentralization efforts almost everywhere else in the post-communist world and which was fueled by a romantic, “small is beautiful” conception of local democracy (Swianiewicz 2002; Illner 1998).

Poland’s reformers argued that communism had failed because the state had tried to do too much. But they did not see the state as inherently leviathan. On the contrary, they insisted that local governments are elements of a single system of public administration “acting within the political, economic, and social structure of the state… and whose range of rights and obligations are determined by law” (Kulesza 1993 p. 40). For them, the key point was that for the national government to fulfill its strategic, policy, and legislative functions, it had to be relieved of the burden of delivering day-to-day public services. A report from the Plenipotentiary for Local Government put it this way:

The excessive concentration of power and competencies at the center creates a situation in which the center continues to be focused on arranging or fixing small things, and not on preparing a broader developmental strategy or on realizing its own policies… [Decentralization is necessary] to increase the responsiveness (sterowalności) of the state and above all the effectiveness of its executive powers (Pełnomocnik 1993 p. 107-8).

In short, the architects of Poland’s reforms saw decentralization as part of a broader state-building strategy and more in functional terms than ones of empowerment.⁹ But they also had little use for fiscal federalism’s tax-based logic of accountability. To be sure, they wanted to give local governments some tax powers as well as large amounts of “lord-less” (bezpanski) state property that they expected would generate revenue. But they knew there was no tax base they could assign to local governments that would yield anything like the revenues needed to fund the functions they wanted to give them. Thus, even at the Round Table, they argued that it was impossible “…to count on local governments to finance themselves from own revenues… [and that] they will require a general grant defined by law and allocated in accordance with objective criteria” (Kulesza 1993 p. 53).

Revealingly, they also rejected a communist proposal to start the decentralization process at the regional level. They argued:

The creation of self-governments requires the existence of a pluralistic form of local social and political life. As we can commonly agree, there still doesn’t exist an appropriately shaped “pro-local government citizens attitude” even at the local level. Self-governments have to be built from the bottom up. Only as a consequence of forming elites (as well as programs) in cities and communes will it be possible to take the next step and build institutions of local government for larger territorial areas (Kulesza 1993 p. 47, italics mine).

Indeed, this statement proved remarkably prescient, accurately anticipating both citizens’ limited engagement with local governments and the centrality that new elites would play in their success. The first local elections of 1990 nicely illustrate both tendencies: Solidarity’s Local Citizens’ Committees nominated the vast majority of the 150,000 candidates who competed for 52,000 council seats. These candidates won a remarkable 80 % of all positions in Large Cities and 47 % nationwide. But while the prospects of running Gmina proved attractive to union members and sympathizers, it failed to excite the average citizen: only 42 % of the electorate voted, 20 % lower than in the parliamentary elections held 9 months earlier (Zespol 1994). Worse, turnout fell to 34 % in the 1994 local elections and has never exceeded 50 %. Moreover, while poll data suggests that most citizens support decentralization, consider local officials more trustworthy than national ones, and think that local governments have improved public services, they do not want to get involved in local affairs—an attitude that Swianiewicz nicely characterized as one of “sympathetic disengagement” (Swianiewicz 2001).

Professionalizing New Elites and Embedding Voice

As Regulski later put it, Poland’s reforms “were initiated ‘from above’ but in the belief—as the future would show to be correct—that local government would become a significant political force” (2000, p. 404). To ensure that they would become a significant political force, Poland’s reformers very deliberately created institutions to train, organize, and give them voice. And remarkably, they started building these institutions even before they created Gmina.

Thus in late 1989, Regulski and a group of MPs established the Foundation in Support of Local Democracy (FSLD) in order to provide training, research, and support services to local governments. The Foundation immediately set up regional training centers which by year’s end had given short courses on the rights and obligations of Gmina to 30,000 (60 %) freshly minted local officials. The Foundation’s centers quickly become research and policy hubs for what activists referred to as the “Local Self-Government Movement” (Ruch Samorządowy, FRDL 2009). By 2009, over a million local officials had been trained by them, while over 10,000 students had earned degrees from the colleges the Foundation eventually created.

Similarly, in early 1990, a team around Kulesza established Municipium, a publishing house devoted solely to local governments. It published the first edition of its national weekly Wspólnota (Community) before the 1990 elections. Wspólnota’s bread-and-butter is explaining new legislation, highlighting intergovernmental disputes, and touting local innovation. But it also reports on local incompetence and malfeasance and has played an important role in holding officials accountable and developing new behavioral norms. In 1991, Municipium launched the academic monthly “Samorząd Terytorialny” (Local Self-Government), a venture that was followed by publication of three more specialized journals on local finances (Finanse Komunalne, 1993), housing cooperatives (Wspólnota Mieszkaniowa, 1996), and human resources (Pracownik Samorządowy, 1998).

In short, Poland’s reformers rather uniquely built institutions designed to knit new local elites into a professional community with a common understanding of its purpose, powers, and problems even before establishing Gmina. Similarly, they sought to institutionally guarantee that this community was integrated into the regulatory regime that would govern the rights, resources, and behavior of local governments.

Thus, provisions in the initial drafts of the Local Government Law (LGL) required the national government to discuss its policies with a single, compulsory Local Government Association. When, however, Parliament rejected a compulsory association,¹⁰ reformers regrouped and introduced articles that established Regional Assemblies of Gmina at the voivodship level (Regulski 2000 pp. 288–90) while also compelling the government to present draft legislation to the local government associations that were established.¹¹

These “little parliaments” or Sejmiki were given substantial regulatory powers over both their constituent members—the Gmina within their voivodships—as well as over the Voivodes who headed the state’s regional offices. The Sejmiki were empowered to conduct independent investigations of professional misconduct by local officials, to intervene in conflicts between mayors and city councils, and to mediate inter-Gmina disputes. Here, they are best understood as centrally mandated, self-monitoring institutions designed to resolve conflicts and improve standards without intervention by the courts.

The Sejmiki were also given powers to oversee the behavior of Voivode, including issuing opinions on their nominees and their performance, and appealing some decisions to the courts. Indeed, the Sejmiki could overturn two types of decision without recourse to the courts: decisions which dissolved municipal councils for alleged financial malfeasance or which prevented the sale of municipal assets (Sejm 1990). The Sejmiki accelerated the transfer of state property to Gmina, monitored the allocation of earmarked grants, helped negotiate agreements between Gmina over the division or joint-management of utilities, and intervened in disputes between mayors and councils—helping to create new norms around the problematic fault line between local executive and legislative powers (Buczkowski 1996). They also nominated half the governing boards of the Regional Accounting Offices, an institution which we will discuss in the following pages.

At the same time, reformers continued in their efforts to ensure that national policies affecting local governments had to be discussed with the associations. In May 1993, the Prime Minister issued an ordinance creating the Joint Commission for Intergovernmental Affairs. The Commission was composed in equal parts of representatives of the national government and the Associations and was empowered to give non-binding opinions on all legislation affecting the finances or functions of local governments (Regulski 2000; Kulesza 2008). Initially, the Commission met sporadically and had limited influence. But when the Sejmiki were eliminated with the Second Phase of Decentralization in 1999, its position was strengthened, first by another ordinance and then by legislation which clarified its structure and powers (Sejm 2005).

A permanent Secretariat was established in the Ministry of the Interior and tasked with providing the Commission’s 24 members with the data, policy research, and draft legislation necessary for its deliberations.¹² By law, the Commission convenes every 2 months and has permanent subcommittees on Finance, Regional Development, and European Integration. In practice, it meets monthly and its subcommittees are in near continuous session. Most importantly, all draft legislation that has any bearing on local governments—now just about everything—must be reviewed by the Commission before it can be submitted to parliament for a vote.

Often, the opinions of the Commission are divided. And even when it reaches a consensus, parliament is free to ignore its recommendations. Nonetheless, the Commission is now involved in all policy making, and its opinions have led to important adjustments in tax sharing; debt regulation; environmental standards; the use of EU funds; the assignment of responsibilities in education, health, transport, and social welfare; and the rules governing accounting, budgeting, and financial reporting. Indeed, as one key architect of the reform noted, “it’s now hard to imagine the operation of the Polish state without it” (Stec 2009).

Embedding of Financial Oversight

Ensuring the fiscal probity of subnational governments is always a challenge. One way to meet this challenge is to make local governments finance themselves, placing the regulatory burden on taxpayers and creditors. Another is to make a government agency or a specialized court responsible for financial oversight. Poland’s reformers never seriously considered the first because they knew local governments would be heavily dependent on transfers. But they also rejected the second because they were afraid that a national government agency might use financial oversight to meddle in local politics. Indeed, they understood that local finance would be a new domain for both Gmina and whoever was tasked with monitoring them, and that working out acceptable practices would be an uncertain process. These uncertainties, they maintained, also argued against arms-length oversight by specialized courts whose well-intentioned ignorance might nonetheless incline them toward overreach (Stępień 1991).

To square this circle, the Poles invented a financial oversight institution all their own—the Regional Accounting Chambers (RIOs). RIOs were first mentioned at the Roundtable Talks (Kulesza, 1992) and included the 1990 LGL. But it was only in 1992 that 17 RIOs were actually created. They were made independent of all line ministries and their boundaries extended across regions, making their jurisdictions different from all other government institutions. Each RIO operated independently, though their directors belonged to a Collegium designed to coordinate their activities. Most importantly, while the Prime Minister nominated their Directors, nominees had to be approved by the Sejmiki, who also named half their managing Boards (Sejm 1992). The result was a decentralized, co-governed institution whose “hybrid”¹³ structure and operational independence were explicitly seen as ways to prevent state overreach while fostering workable financial standards (Dębowska-Romanowska 2013).

In 1993, seven RIO directors came from Gmina, four from the inner circle of reformers, two from state inspectorates, and three from other professions. Almost all of the 98 Board Members appointed by the Sejmiki (of 196) worked in local governments. The Boards then recruited close to one thousand employees to be inspectors and analysts (Stec 2004). The vast majority of these were recent college graduates or accountants and lawyers drawn from industry because, as one RIO director put it:

We were convinced that we couldn’t hire people with habits carried over from the old institutions of control because local self-government was a completely new systemic solution that has its own, quite different legal foundation than other parts of public administration. (Cybulski, Stec, Wspólnota 27/06/13)

The RIOs responsibility for overseeing the “financial economy of local governments” is divided into four functions: Control, Oversight, Opinion Giving, and Information and Training. Under “Control,” they can review the legality—not substance—of all local budget resolutions. They can declare resolutions illegal and even take temporary control of a municipality’s finances if its Council fails to comply with the law. Under “Oversight,” they must audit the finances of all local governments once every 4 years and they must issue non-binding opinions—“Opinion Giving”—on a local government’s ability to service prospective debts. Finally, they are responsible for developing annual training programs for their jurisdictions (Sejm 1992) because reformers thought training was critical for the creation and dissemination of new norms (RIO 2006).

Unfortunately, there are no detailed studies of the RIOs during their initial years of operation, but data on their rulings is suggestive. Table 1 shows the number of budget resolutions reviewed by the RIOs for select years between 1993 and 2011, as well as the share of resolutions they declared in violation of the law; the share of those violations which they considered insignificant; the share of all violations in which they initiated procedures to declare a resolution invalid; and the share of all violations that local governments failed to correct and which the RIOs ultimately invalidated.

Table 1 Budget resolutions reviewed by the RIOs for select years between 1993 and 2011

In the early 1990s, the RIOs found a high proportion of budget resolutions in violation of the law. But until 1998, they chose to consider the majority of these insignificant. They also initiated corrective procedures in a relatively small number of cases, either because they were cutting Gmina slack or because most decided to correct the problems themselves. Finally, they declared a very small percentage of significant violations invalid, probably because most Gmina ultimately choose to change their budget resolutions themselves. So in the early years of reform, when local financial management was shakiest, the RIOs pursued a policy of encouraging compliance through reprimands, instruction, and self-correction, rather than sanctions. Or, as one director put it, “We considered prevention more valuable than imposing consequences on local governments who broke the law” (Cybulski and Stec 2013).

But over time, the policy shifted. As local governments learned the ropes, the share of resolutions found in violation of the law fell from 27 % to under 5 %. This change occurred despite a sharp increase in the total number of resolutions reviewed and suggests that the RIOs pedagogic mission was being achieved. At the same time, their rulings hardened between the 1990s and the 2000s: They considered more violations serious (c. 25 vs. 65 %), they subjected more to corrective procedures (20 vs. 25 %), and ultimately declared a much higher percentage of them illegal (c. 10 vs. c. 35 %).

This shift in policy seems to be due in part to the fact that more of the (fewer) cases that the RIOs considered in violation of the law were really serious; in part, because the RIOs have become more rigid over time,¹⁴ and in part because there has been an increase in local governments who refuse to correct their resolutions because they want to challenge RIO rulings in the courts. Nonetheless, by embedding local governments into the management structures of the institutions created to monitor them, and by endowing these “hybrid institutions” with wide discretionary powers and training functions, Poland’s reformers succeeded in institutionalizing the development and adoption of prudent financial norms and practices at the local level—no mean feat.

Education Decentralization as an Embedded Process

The literature on decentralization maintains that local governments work best when their financial and managerial obligations for public services are clearly defined: Without clearly separating responsibilities across levels of government, the argument runs, neither citizens nor oversight institutions can reasonably hold actors accountable. Thus, even when a responsibility is defined as “shared,” policy makers should try to legally parse its composite tasks as neatly as possible between levels of government (Musgrave 1959; Ahmad et. al. in Ter Minassian 1997). This sounds sensible enough. But it has little to do with how education functions were divvied-up in Poland.

In the heady days of 1990, Poland’s reformers legally defined both preschool and primary education as “public functions that Gmina execute in their own name and on their own responsibility” (Sejm 1990, italics mine). Nonetheless, the meaning of “own” was—and remains—very unclear. One major problem was created by the teachers’ unions, who forced reformers to maintain the national government’s control over wages and working conditions. Thus, Gmina finance and manage schools in their “own name and responsibility,” but teachers wages are set by parliament (Levitas and Herczyński 2002).

Reformers themselves created another problem: In 1990, they required Gmina to take over preschools and fully finance them from their general revenues. From a fiscal federalist point of view, this made sense because preschool education had been legally defined as an “own responsibility.” But the law also defined primary education as an own responsibility. Here, however, not only were Gmina given a few years to assume control over primary schools, they were also given grants to help pay for running them. The purpose of these grants, however, was never specified, and people simply assumed that they would cover the operating cost of primary schools, while Gmina would pay for investments from other sources (Herbst, Herczyński, and Levitas 2009).

From a fiscal federalist point of view, these arrangements made no sense: If education is a local government “own responsibility” but the national government sets teacher’s wages, then there is no legal obstacle to the national government jacking up wages and leaving local governments to pick-up the tab. Conversely, the failure to define which costs the education grant was designed to cover made it possible for local governments to hire teachers—or not fire those made redundant by declining enrollment—and then claim that the national government should foot the bill. Indeed, both problems have haunted education finance since the early 1990s. Nonetheless, neither has prevented the Poles from radically improving their schools.

The basic reason for this success lies in the institutional architecture that reformers built to ensure that local governments were always involved in national policy making. This architecture made it possible for the Poles to reach a series of pragmatic compromises that rendered the incoherence of the legal regime if not moot, fungible. And while it is impossible here to trace the politics behind the compromises, it is useful to sketch how the problems associated with both the education grant and national wage setting played out in practice.

In the early 1990s, when Gmina began to take over primary schools, the national government gave them grants equal to what it had previously spent on them. But as the number of Gmina who took over their schools rose, this practice became untenable; partly because it became clear that the historical allocation of funds was wildly unequal, and partly because demography and migration were changing enrollment patterns. Thus, an agreement was reached between the Ministry and the local government associations to develop a formula which would allow “money to follow pupils.”

This formula could be developed around a “bottom-up” calculation of what education should cost per pupil given certain assumptions about inputs and standards. Or it could be driven by a “top-down” calculation that assumed that existing spending was close to what the country could afford, and that a reasonable approximation of per pupil costs could be had by simply dividing current spending by the total number of pupils (Levitas and Herczyński 2002). Not surprisingly, everybody preferred the “bottom-up” strategy. But after repeated efforts to estimate what education should cost yielded sums well above what was being spent, the Ministry opted for a “top-down” approach.

The simple calculation of average per pupil costs, however, was insufficient to create a workable grant system because different pupils (e.g., the handicapped, minorities) in different jurisdictions (e.g., rural and urban) had different costs. To accommodate these differences, the Ministry assigned coefficients to different types of pupils in different types of jurisdictions. The most important coefficient provided a third more money for pupils in rural schools on the grounds that classes in these schools were, on average, 30 % smaller than in their urban counterparts. Nonetheless, despite this additional funding, many Rural Gmina had trouble maintaining small schools.

For our purposes, however, what is important to understand is that over time it became impossible to tell whether any gap between a Gmina’s actual education spending and what it received through the grant was due to its own behavior—e.g., failure to close facilities and/or reduce employment—or to the coefficients used in the formula (Herbst, Herczyński and Levitas 2009). And as the fiscal federalists would anticipate, the failure to specify what the education grant is really supposed to cover has generated an endless-blame game: Local governments argue that the grant is insufficient to cover the basic costs of schooling, while the national government insists that it is up to local governments to make due with whatever they get because education is an “own responsibility.”

In practice, however, the institutional architecture for intergovernmental dialogue created by reformers has allowed these ostensibly irreconcilable positions to be bridged through a series of pragmatic compromises and adjustments. For starters, the legal regime requires the Ministry of Education to discuss with the local government associations—and now the Joint Commission—both its annual budget request to parliament and a draft of the formula it will use to allocate the education component of the general grant. The discussion of the coefficients always produces tensions between the local governments because of their distributional consequences. At the same time, everybody also understands that some of these tensions can be resolved by increasing the Ministry’s overall budget and with it the size of the grant (Levitas 2012).

It is impossible to trace here the politics behind either changing coefficients or the size of the grant pool, but three general points can be made: First, local governments have been instrumental in defending (or expanding) the Ministry’s budget for primary and secondary education, and in ensuring that the demographic pressure to close schools has not been compounded by budget cuts. Second, during the 1990s, the coefficients favored Rural Gmina because people understood that it was harder for them to adjust their school networks to falling enrollment. But this changed in the 2000s as demographic decline washed through the higher grades, forcing Large Cities to close high schools. And third, the debates over both the budget and the formula became increasingly sophisticated, with all sides deploying better data and studies to justify their positions (Herbst, Herczyński, and Levitas 2009).

These same institutional arrangements have been used to manage the contradiction between central wage-setting and local wage-paying. Here too, the Joint Commission has forced the Ministry to make incremental adjustments in the education budget to cover wage increases and/or negotiate changes in teachers’ benefits that effect local budgets. But the best proof of the importance of these institutional arrangements is how they were used to overcome precisely the kind of catastrophic failure that the fiscal federalists would anticipate given the contradictions in the legal regime.

This failure came in 1999, as the Ministry was working on two distinct reforms associated with completing the Second Phase of Decentralization: First, it was rewriting the per pupil formula in order to accommodate the decentralization of secondary education. And second, it was redesigning the wage system in order to create stronger incentives for professional advancement. But the reform efforts were not coordinated and the department responsible for the new wage system underestimated the number of teachers that would be immediately entitled to raises. As a result, local wage costs increased dramatically over what the grant was designed to finance, quickly putting many local governments on the edge of insolvency. Worse, the problem required more than adding money to the grant pool because the teachers entitled to increases were unevenly distributed across jurisdictions. Finally, and further complicating matters, some local governments clearly overstated the number of teachers entitled to wage increases in order to maximize any future settlement (Levitas and Herczyński 2002).

To resolve the crisis, the government called a series of meetings with the Joint Commission and the RIOs. The RIOs were mobilized to verify the number of teachers entitled to immediate wage increases and to estimate the additional wage costs of each jurisdiction. The Minister of Education was forced to resign and coefficients were added to the formula to adjust grants in accordance with the percentage of teachers at different pay grades in each jurisdiction. But as before, the entire mechanism remained unanchored by a clear legal definition of the financial responsibilities of different levels of government. Thus, the resolution of the crisis did not fix the problem that precipitated it. Nonetheless, Poland’s framework for, and habits of intergovernmental negotiation made possible a workable compromise (Levitas 2012).

What Does the Polish Case Tell Us About Decentralization

Poland is now one of the more decentralized countries in Europe. Its subnational governments have transformed the nation’s infrastructure and improved its schools. This success cannot be attributed to the rules fiscal federalists consider critical for enhancing good local governance. Despite their “transfer dependency” and the profound confusion surrounding some of their key responsibilities, Polish local governments have “delivered the goods.” Nor can this success by explained by popular demands for local control or by an engaged citizenry playing a particularly active role in holding newly elected officials accountable. On the contrary, Polish civil society appears to be as weak and disaffected as its counterparts in the rest of post-communist Europe. In short, Poland’s success defies the expectations of the dominant discourses about decentralization.

Instead, decentralization in Poland was “a revolution from above,” pushed forward by a core group of reformers who saw decentralization more in terms of public administration reform than democratic empowerment, and whose principle goal was to build a stronger state. Skeptical about their co-patriots readiness to actively participate in their own governance, and conscious that the transformation of the single-party state would be a long struggle, they placed their bets on the local officials that their reforms would create.

That these bets paid off is due in no small part to the fact that—like the architects themselves—many of these officials came from the ranks of the civic-minded and politically engaged elites that had emerged during Poland’s long struggle against communism. But recognizing the importance of these elites is very different from arguing that the success of the reforms as a whole should be attributed to a particularly robust civil society. Instead, Poland resembles other cases in which particular social groups come to play an outsized role in state building and development (Anderson 1983; Skoczpol, Evans, and Rueschemeyer 1985; Evans 1995; Petro 2001).

Indeed, the more interesting question is how these elites were forged into an instrument of effective governance. My answer to this question is that the architects of the reforms very consciously created an array of institutions designed to train, professionalize, discipline, and empower them. Horizontally, these institutions molded newly elected local officials into a professional group that shared a common language and purpose. They transformed local government reform into something of “movement” while disciplining this movement through monitoring mechanisms and the collaborative development of new norms and standards. Vertically, these institutions embedded local governments in the regulatory structure of the state, allowing local governments to represent themselves while facilitating the continual adjustment of the powers, responsibilities, and finances of different levels of governments.

In part, these institutions functioned well because they combined the cultivation of a common sense of purpose with the development of new norms and standards. In part, because they encouraged the continual adjustment of the intergovernmental system as a system. The former is in line with much of the work that highlights the centrality of shared goals, identities, and norms to successful organizations and public institutions (Selznik 1957; Passoti 2010). The latter has affinities with the literature on pragmatism and experimental governance, a literature which identifies self-monitoring and the capacity for continuous adjustment of rules and practices as central to better government (Ansell 2011).

How well these institutions might translate to other contexts is obviously uncertain. But it does seem that if the Polish case tells us anything, it is that successful decentralization—as state building—requires an array of overlapping institutions to train, monitor, professionalize, and represent local governments in a long, evolutionary process. This process, and the institutions needed to negotiate it, belie the fiscal federalist project of imposing a clear set of rules whose tax-based logic of accountability are held to ensure good governance.

But the Polish case also demonstrates that impressive gains in governance can be made without a particularly engaged citizenry. Saying this, of course, is not an argument against engaging citizens in their own governance. Nor is it to say that there are not many contexts in which mass mobilization and popular contention are the best tools for creating more responsive government, local or otherwise. Nonetheless, Poland’s success suggests that if there has been a political breakthrough at the national level, then the construction of effective, responsive, and responsible local governments may be better pursued by focusing on the institutions necessary to train, professionalize, discipline, and empower newly elected local elites, than by looking first to the direct participation of citizens.