The quality of government and administrative performance: explaining Cohesion Policy compliance, absorption and achievements across EU regions

ABSTRACT

Does the quality of government improve the administrative performance of regions in European Union (EU) Cohesion Policy? This article analyses the relationship between the quality of government and multiple dimensions of administrative performance in Cohesion Policy. Using primary data on government quality combined with EU performance data for the European Regional Development Fund (ERDF) in the period 2007–13, regression analysis is undertaken for 173 European regions. The results confirm that government quality is a key determinant of administrative performance in terms of financial compliance, timely spending and outcomes. The findings support the need for capacity-building in regions with low quality of government to improve policy implementation.

KEYWORDS:

• European Union Cohesion Policy
• quality of government
• administrative capacity
• regional development

JEL:

• O43
• O52
• R8

1. INTRODUCTION

The importance of subnational quality of government for public sector performance is increasingly appreciated by policymakers and scholars, but remains understudied. The literature on federalism and intergovernmental relations in the US context has long recognized that the capacity of subnational governments is critical for good performance in federal, state and local authorities, and for the effectiveness of these jurisdictions in accessing and spending intergovernmental grants (reviewed by Carley et al., 2015; Nicholson-Crotty, 2015). Similarly, the Organisation for Economic Cooperation and Development (OECD) has highlighted the critical role of subnational capacity for the effective delivery of public investment programmes and grants, given that economic development policies are usually administered by subnational governments and because capacity is critical for improving policy coordination, strategic planning, project selection, expertise and learning (OECD, 2013, 2014, 2016, 2018).

In the European Union (EU) context there is an extensive literature investigating the implementation of Cohesion Policy – the main intergovernmental investment instrument – and the implications for wider territorial relations and subnational empowerment (Bachtler & Mendez, 2020a). Much less attention has been placed on how subnational capacity impacts the delivery of formal policy objectives, although the importance of this link is increasingly recognized by EU institutions (European Commission, 2014a, 2017a, 2018, 2022; European Committee of the Regions, 2018). Studies have found that the quality of regional government impacts regional business and economic performance (Bachtrögler & Oberhofer, 2018; Becker et al., 2013; Muringani et al., 2019; Rodríguez-Pose, 2013; Rodríguez-Pose & Di Cataldo, 2015; Rodríguez-Pose & Garcilazo, 2015; Rodríguez-Pose & Ketterer, 2020) and accessing and using EU funding (Crescenzi & Giua, 2016). However, while there are several national studies of spending or compliance (e.g., Cace et al., 2009; Hagemann, 2019; Mendez & Bachtler, 2017; Surubaru, 2017; Tiganasu et al., 2018; Tosun, 2014), the role of quality of government in the administrative performance across EU regions has received less attention, and is often restricted to a limited number of national or regional case studies (e.g., Baun & Marek, 2017; Milio, 2007; Terracciano & Graziano, 2016).

The main research question of this article is whether the quality of government influences the administrative performance of Cohesion Policy across EU regions. Cohesion Policy is a crucial case for assessing administrative performance at the regional level for several reasons. First, it is the EU’s main investment programme, accounting for 37% of the EU budget in the period 2014–20 with a budget of around €350 billion, and a similar allocation for 2021–27. Second, there is now comparative administrative data available for assessing administrative performance. Third, the policy is implemented under shared management, with a strong role for regions in policy delivery and implementation, a necessary condition for assessing regional variations in administrative performance and governmental quality. Although the Commission remains responsible for the implementation of the EU budget, the management, implementation and control of EU funds and programmes are delegated to national and regional authorities, which select beneficiaries, distribute funds, oversee implementation according to EU rules and ensure performance against agreed targets.

EU Cohesion Policy is also salient because it has come under greater scrutiny since the economic crisis, particularly in debates on the 2021–27 EU budget and the future of the EU. Contributions to the debate on the post-2020 Multiannual Financial Framework were critical of the added value and performance of Cohesion Policy (reviewed in Bachtler et al., 2016a, 2017). The Commission’s White Paper on the Future of Europe even identified a future scenario – ‘doing less more efficiently’ – involving the termination of or reduced funding for Cohesion Policy due to perceptions of it having limited added value or being unable to deliver on its promises (European Commission, 2017b, p. 22).

In order to understand how well the implementation of the funds is being administered, this analysis focuses on three key objectives of administrative performance: timely absorption of funding (policy inputs); the legality and regularity of expenditure (policy management); and achievement of outcome targets (policy outputs¹). These three dimensions of administrative performance have been selected for analysis because they constitute the key goals of Cohesion Policy implementation and are used by the policy’s principals (at EU and national levels) to account for the effective management of Cohesion Policy resources, as well as featuring frequently in debates on Cohesion Policy in the EU Parliament, European Committee of the Regions and media (Mendez et al., 2020). Previous research has tended to focus only on the first of these (absorption), much less on the other two measures, and has never examined the measures in combination. Analysing these three dimensions also provides insights whether there are trade-offs between timely spending, compliant spending and effective spending, as some of the literature on conditionalities has suggested (Bachtler & Ferry, 2015; Bachtler & Mendez, 2020b). Further, there are robust and comparative administrative data available for assessing these dimensions of performance for a complete programme period, 2007–13. Lastly, understanding whether and how government quality leads to implementation performance failure or success in Cohesion Policy sheds light on potential management or support techniques that EU and national institutions can use to minimize this failure.

Absorption of funding and achievement of the outcome targets are prerequisites for achieving long-term impacts of EU funding, and compliance with legal requirements contributes to sound financial management and avoids investment delays arising from sanctions. Moreover, if programme implementation is slow, irregular and does not deliver outcome targets, citizens and the media regard the policy as ineffective even if positive impacts materialize in the future.

Using primary data on government quality collected by the Gothenburg Institute of Quality of Government combined with data on European Regional Development Fund (ERDF) administrative performance in 2007–13, regression models are estimated for 173 European regions to examine the impact of quality of government on three dimensions of administrative performance while controlling for a range of other factors. In doing so, this article makes three key contributions.

First, it provides a systematic analysis of the impact of the quality of government on the administrative performance of Cohesion Policy across EU regions for the first time. Second, it advances a more refined and multidimensional conceptualization of administrative performance in Cohesion Policy by distinguishing the delivery goals of absorption, compliance and achievement of target outcomes. Third, it contributes to the academic and policy debates on ‘good governance’ and capacity-building to support effective delivery in Cohesion Policy, addressing the so far limited conceptual or empirical analysis of the different dimensions of administrative performance and the link to regional capacity.

The remainder of the article is structured as follows. It begins by reviewing the theoretical and empirical literature on administrative performance and quality of government. The research design, methodology and data are then presented. The empirical results and analysis follow, and the final section concludes and sets out the key theoretical and policy implications.

2. ADMINISTRATIVE PERFORMANCE AND QUALITY OF GOVERNMENT

The literature on federalism and intergovernmental relations has long recognized that the capacity of subnational governments is critical for good performance (Derthick, 1970; Elazar, 1984; Gamkhar & Pickerill, 2012; Hall, 2012; Jennings et al., 2012). Capacity has been shown to correlate with implementation performance in national, regional and local authorities (e.g., May, 1993; McDermott, 2006; Spillane & Thompson, 1997). It has also been associated with the effectiveness of different jurisdictions in winning and spending intergovernmental grants (e.g., Collins & Gerber, 2008; Hall, 2008; Handley, 2008; Manna & Ryan, 2011; Terman et al., 2016; Terman & Feiock, 2015).

The impact of institutional quality on the economic and administrative performance of EU investment policies has received increased academic and policymaker attention in recent years, particularly in EU Cohesion Policy. The European Commission’s successive Cohesion Reports have dedicated chapters to the importance of good governance and institutional improvement and highlighted how poor governance can reduce Cohesion Policy’s economic impact and leverage effects (European Commission, 2014a, 2017a, 2022), supported by the analysis of quality of governance data sets from the World Bank (world governance indicators – WGI), OECD (sustainable governance indicators – SGI), Transparency International and Gothenburg University (Quality of Government Expert Survey).

The Cohesion Reports also drew on evidence from econometric studies, which have found that the quality of institutions have both a direct impact on economic performance and a moderating effect on the impact of Cohesion Policy on growth across EU member states (Ederveen et al., 2002, 2006), especially on less-developed EU regions (Becker et al., 2013; Fazekas & Czibik, 2021; Rodríguez-Pose & Garcilazo, 2015; Rodríguez-Pose & Ketterer, 2020). A further Commission study concluded that ‘lack of human capital and poor institutional quality hampers competitiveness and investment decisions’, one of five sets of factors explaining weak growth in low-income and low-growth regions of the EU and requiring investment to strengthen institutional capacity and the efficiency of public administrations (European Commission, 2017c, pp. iv–v).

On the administrative side, slow rates of financial absorption and other implementation difficulties are often attributed to weaknesses in the quality of government (European Commission, 2014a, 2017a). These include a lack of competence or insufficient staff; weaknesses in management systems; poor coordination between different bodies; incorrect implementation of public procurement, environmental and state aid rules; and patchy use of performance management and anti-corruption techniques. They are also attributed to politicization, rent-seeking, clientelism, corruption and other irregular use of EU funding (Fazekas, 2017; Hageman, 2019a).

The importance of administrative capacity for sound programme management is emphasized in EU-wide evaluations of Cohesion Policy (Bachtler et al., 2009, 2016b; KPMG & PROGNOS, 2016) and strategic policy reviews (Barca, 2009). The ex-post evaluation of the delivery systems of the 2007–13 programmes highlighted the importance of administrative capacity for effective programme delivery. Gaps in technical knowledge and capacity in managing authorities and intermediate bodies constituted a challenge for the legality and regularity of expenditure and undermined compliance (KPMG & PROGNOS, 2016, p. 104). Timely spending was dependent on ‘the maturity and capability of the institutional system’ responsible for the implementation of the programmes (p. 58), although there was limited verifiable evidence or data provided for these conclusions.

Academic studies have found a positive link between administrative capacity and financial absorption based on small-n comparative studies of two regions in Italy (Milio, 2007; Terracciano & Graziano, 2016) and different combinations of Central and Eastern European member states (Bachtler et al., 2014; Baun & Marek, 2017; Hagemann, 2019a, 2019b; Surubaru, 2017; Tiganasu et al., 2018). While the primary focus of the studies on Central and Eastern Europe has been on the national level, the qualitative study on the Czech case suggested ‘the primary importance of regional administrative capacity and performance as a factor affecting Cohesion Policy implementation’ (Baun & Marek, 2017, p. 863).

Quantitative studies across all EU member states have come to similar conclusions. Statistical analyses of the impact of quality of government on financial absorption (Incaltarau et al., 2019; Tosun, 2014) and financial compliance (Mendez & Bachtler, 2017) have found a statistically significant and positive effect. These studies are restricted to the national level (rather than regional level), limiting the number of factors that can be controlled for given the small number of observations. The only study to have examined financial absorption across EU regions is by Kersan-Škabić and Tijanić (2017). While a significant and positive correlation was found between absorption and the quality of government, the study did not analyse the standard spending rate measure used to measure absorption by EU and national institutions (i.e., funds spent as a share of total programme funding). Instead, the study analysed spending per head of the population (i.e., financial intensity); and spending as a share of funding committed, which can diverge significantly from the allocation to programmes (due to differences in the temporal profile of commitments across programmes) and does not therefore accurately measure the spending performance of a programme comparatively.

Going beyond these studies, the core theoretical hypothesis investigated in this article is that the quality of regional government (as a proxy of administrative capacity) is positively associated with the administrative performance of regional programmes under EU Cohesion Policy. We adopt a multidimensional conceptualization of administrative performance distinguishing compliance, spending and achievements in order to provide a more comprehensive assessment than previous studies. In the following section we set out the research design, methodology and data used to investigate the relationship between administrative performance and the quality of subnational government.

3. RESEARCH DESIGN AND DATA

To test the effect of the quality of government on Cohesion Policy implementation performance, a unique dataset was compiled using cross-sectional data at the regional level based on a range of EU sources. The geographical coverage of the regional sample comprises 172 regions in 17 member states. The sample excludes regions where the regional data do not correspond to the programme’s geographical area. For instance, the four Dutch regional operational programmes (OPs) cannot be included because they correspond to NUTS-2 regions, whereas the quality of government indicator covers 12 smaller NUTS-3 regions that cannot be aggregated. For a full list of the NUTS-1/2 regions in the 17 member states included in the sample, see Appendix A in the supplemental data online.

The programming period analysed is the 2007–13 period and specifically the final year of the period (2013). This year provides a useful benchmark for assessing implementation performance over the full course of a programme cycle because it is the final year in which funds can be committed to projects (see also Tosun, 2014). It is also the only year for which there are regional data available for the administrative performance indicator of compliance (error rate).

3.1. Dependent variable: administrative performance

The administrative performance of Cohesion Policy is multidimensional. As noted, we distinguish three dimensions relating to: timely spending; financial compliance; and the achievement of outcome targets. These are the main dimensions of administrative performance used by EU institutions to oversee and account for performance in: the annual reports on EU budget implementation issued by DG Budget; the budgetary discharge procedure on the legality and regularity of EU spending (involving the Council/European Parliament, drawing on the European Court of Auditor’s annual reports); and the annual activity reports on performance in achieving targets (set in the Commission’s strategic/management plans). Accordingly, the three dimensions can be measured directly using administrative data that are reported to the EU.

Financial absorption is the first dimension of administrative performance. The absorption rate measures the spending performance of programmes and is defined as the percentage of available funding paid out by the European Commission to the 2007–13 programmes relative to the total financial allocation of each regional programme. Across the regions covered in this sample, the financial absorption rate varies between 65% and 95%, with higher values indicating stronger financial implementation performance.

The second dimension of administrative performance measures financial compliance in terms of the regularity and legality of spending. Anchored in EU treaty and financial regulation commitments to protect EU financial interests from administrative errors in financial management, such as claiming for ineligible expenditure, the measure of financial non-compliance used is the net ‘error rate’ (technically known as the cumulative residual risk rate – CRR). It is an estimate by the European Commission of the part of the expenditure declared, for each programme during the entire programme period, which is not considered legal and regular.² The CRR takes account of all financial corrections implemented since the start of the period and is expressed as a cumulative percentage of the value of the payments made to the programme which are not in conformity with EU rules, with lower (higher) values indicating greater compliance (noncompliance).³

To ease the interpretation and comparability with the two other performance measures used in the analysis, the scale has been reversed so that higher values indicate better compliance performance (instead of non-compliance) and lower values correspond to lower compliance.

The achievement of outcome objectives is the final administrative performance indicator. This measures quantified output and result indicators (such as the number of new jobs created, small and medium-sized enterprises (SMEs) assisted, kilometres of roads constructed, etc.) that contribute to programme objectives and are assessed according predefined targets. Outputs are accomplished with the resources allocated to an intervention (e.g., training courses delivered to unemployed young people, number of sewage plants or kilometres of roads built, etc.). Results are a measurable consequence deriving – directly or indirectly – from a cause-and-effect relationship. A dataset of these core indicators across programmes was constructed as part of the ex-post evaluation of Cohesion Policy based on data collected from the annual implementation reports by member states.⁴ Achievement of targets are calculated by dividing the 2013 target value by the 2013 achievement value for all programme indicators and then providing an average programme rate. We use this method because it replicates the approach agreed by the EU institutions to measure performance and allocate the performance reserve (European Commission, 2000, 2014b).

3.2. Independent variable: quality of government

The core independent variable of theoretical interest in this study is the quality of regional government. The empirical measure used is the European quality of government index (EQI) (Charron et al., 2014, 2015; Charron & Lapuente, 2013), based on data from two surveys conducted in 2010 and 2013. The 2013 survey was based on responses from 85,000 citizens in 206 regions (at NUTS-1 or -2 levels) covering 24 countries (21 EU member states). The survey asked respondents 16 questions about government quality, impartiality and corruption focussing on three policy areas (education, health services and law enforcement) that tend to be governed or administered at regional level in order to maximize regional variation. The European EQI is formed by aggregating and standardizing the mean scores for each question with mean of 0 and a standard deviation (SD) of 1. The independent variable is therefore a measure of citizens’ perceptions of the regional quality of government.

Figure 1 maps the quality of government by region revealing large variations across the EU. The values range from −2.24 (Campania in Italy, a low quality of government) to 2.78 (Åland in Finland, a high quality of government).

Figure 1. Quality of government variations across European Union regions.

Source: Authors’ own elaboration based on the European quality of government index (EQI).

3.3. Control variables

To avoid finding spurious relationships, it is necessary to include relevant control variables. Informed by existing theoretical and empirical literature on Europeanization and compliance, the control variables relate to regional authority, goodness-of-fit, EU preferences and monitoring, EU membership status and the impact of the economic crisis.

3.3.1. Regional authority

Territorial governance arrangements can influence implementation outcomes by determining the number and power of veto points in the policy process (Haverland, 2000; Mbaye, 2001; Steunenberg, 2006). In political systems that grant subnational/regional governments greater autonomy (e.g., decentralized or federal political systems compared with centralized systems), there is a greater opportunity for regional governments to block compliance, while more centralized systems may be able to enforce compliance centrally. Theoretically, the impact of regional autonomy on Cohesion Policy implementation can also be positive (Bahr, 2008; Tosun, 2014). Political decentralization could facilitate effective Cohesion Policy implementation if it promotes more accountability in policy delivery and may support spending and compliance (Incaltarau et al., 2019; Polverari, 2015; Surubaru, 2017).

To operationalize the extent of a region’s autonomy, the regional autonomy index (RAI) is used (Hooghe et al., 2016). It measures the formal authority of intermediate or regional government combining measures of self-rule (in terms of institutional depth, policy scope, fiscal autonomy, borrowing autonomy and representation) and shared rule (role of regions at the national level law-making, executive control, fiscal control, borrowing control and constitutional change).

Aside from this general measure of political–institutional autonomy, it is also important to consider the role of Cohesion Policy-specific measures of autonomous regional decision-making (Mendez & Bachtler, 2017). The rationale underpinning Cohesion Policy’s multilevel governance model of implementation is that decentralized and partnership-based delivery can facilitate the tailoring of the policy to regional/local contexts and support effective implementation by eliciting local knowledge and increasing ownership (Barca, 2009). Accordingly, we have created a Cohesion Policy regional autonomy index (CPRAI) (see Appendix A in the supplemental data online) to measure the level of regionalization of ERDF programme management authority. Specifically, we identify whether the three key bodies responsible for administering the ERDF are located at regional or national level. The bodies are the managing authority (MA), which has overall operational responsibility for the efficient management and implementation of the programme; the certification authority (CA), which makes payments claims to the Commission and ensures eligibility of rules; and the audit authority (AA), responsible for verifying the effective functioning of the management and control systems. As all three institutions collectively play a critical role in the effective and efficient administration of EU regional funding, we have computed a variable to identify the level of regionalization across the three bodies. The CPRAI is a four-point scale as follows:

• Fully centralized: where all authorities (MA, CA, AA) are located at the national level (e.g., Hungary, Slovakia, Sweden, Spain, the UK (English) regions).

• Regionalized MA: programmes where the MA is at regional level, while the CA and/AA are at a national level (e.g., Poland, Portugal).

• Regionalized MA and AA or CA: programmes where the MA is at a regional level, along with a CA or AA at a regional level (e.g., France).

• Fully regionalized: programmes where all three authorities (MA, CA and AA) are located at the regional level (e.g., Belgium, Germany, devolved UK regions, Åland region in Finland).

3.3.2. Goodness-of-fit

One of the most prominent explanations of the Europeanization of member states policies and patters of compliance is the ‘goodness-of-fit’ between EU and domestic policies and goals (Haverland, 2000), the expectation being that resistance to EU pressures and non-compliance with EU rules increases with the level of ‘misfit’ between the domestic status quo and EU policies or goals (Héritier, 1996). Many studies have confirmed the positive effect on the compliance performance of goodness-of-fit. With respect to Cohesion Policy, Mendez and Bachtler (2017) found quantitative and qualitative evidence for the role of goodness-of-fit between EU and national goals and requirements in explaining financial non-compliance patterns across member states.

We use the ‘co-financing rate’ as a proxy for goodness-of-fit. It measures the contribution of EU funding to a programme expressed as a percentage of the total programme cost. This is based on research evidence that the larger the share of public investment accounted for by Cohesion Policy, the greater the alignment between EU and domestic regional development strategies, instruments and rules – and vice versa (Davies & Polverari, 2015; EPRC & EUROREG, 2011; Mendez & Bachtler, 2017).

3.3.3. EU preferences/financial interests

It would be reasonable to expect that the Commission places more scrutiny on the implementation and performance of regions receiving the greatest volume of funding, because of the greater potential to achieve overall goals as well as the greater financial and reputational risks to the EU budget from poor performance or non-compliance in these regions. There is also pressure from the member states for the Commission to do so, notably from the net payers into the EU budget to ensure that their own contributions are used effectively and in accordance with EU rules in net beneficiary countries (Mendez & Bachtler, 2011; Pollack, 1995). EU financial interests are measured through an indicator on the ERDF funding to each regional programme as a share of the total ERDF funding for all programmes in the EU budget.

3.3.4. Economic factors

The literature highlights that the degree to which member states have been affected by the recession resulting from the economic and financial crisis must be taken into account in assessing implementation performance such as absorption, although the impact on spending is contested (Tosun, 2014; Hagemann, 2019b; Incaltarau et al., 2019). The economic crisis reduced demand for certain measures and created difficulties in raising domestic co-financing for interventions in many member states because of fiscal consolidation pressures (European Commission, 2013). We measure the impact of the crisis using Eurostat data on changes in gross domestic product (GDP) per capita (pps) over the period 2007–11. Given that the EU introduced crisis flexibility measures to facilitate absorption (through reduced co-financing obligations), we also control for programmes that took up this measure with a dummy variable. This control is therefore only relevant to the absorption rate.

3.3.5. EU membership status/experience

Previous studies have highlighted that member states that acceded to the EU in the 2000s (EU-13) were under more political pressure to administer programmes effectively and comply with EU rules. This is because of a need to justify redistributive shifts in funding away from the EU-15 members towards the EU-13 member states and to avoid providing net payer countries with arguments for cutting funding in budget negotiations (Bachtler et al., 2014; Mendez & Bachtler, 2017). A dummy variable is included in the model to distinguish the EU-13 member states.

3.3.6. Type of expenditure

The type of programme expenditure may also impact on performance, given the different characteristics of interventions and projects. In particular, programmes with large infrastructure investments may be more challenging to implement because of the lengthy and complex project planning, assessment and public procurement processes involved (Hagemann, 2019). Indeed, problems in the application of procurement rules for infrastructure projects are a major source of administrative errors and delays implying a higher risk for successful administrative and financial performance (Mendez & Bachtler, 2017). To control for the presence of large-scale infrastructure investments, we include a variable measuring the share of the regional programme accounted for by basic infrastructure investments in the transport, energy and environment sectors.

3.3.7. EU compliance monitoring

Theories of enforcement view compliance as a function of the expected costs from the likelihood of detection and credible threat of sanctions. Rigorous monitoring by EU institutions is expected to increase the costs of non-implementation/compliance (Tallberg, 2002; Zhelyazkova & Torenvlied, 2011). We measure EU compliance monitoring through an indicator on the number of so-called ‘reservations’ issued by DG REGIO following audits of compliance. This control is only modelled for the spending and absorption performance as it is not theoretically relevant to include compliance audits as a predictor of the compliance rate, given that the latter is dependent on the former. The data were supplied to the authors by DG REGIO.

3.3.8. Models

Descriptive statistics for each variable are presented in Table 1. As the outcome variables (compliance, absorption and achievements) are percentage rates bounded at both ends (by 0% and 100%), standard linear/ordinary least squares (OLS) (or Tobit) regression is not an appropriate estimation modelling method because it does not account for the fact that percentages are bounded leading to biased and inefficient results. The most appropriate estimation method is Beta regression because it models the distribution of the dependent variable using the beta distribution, which is bounded between 0 and 100 or 0 and 1 when transformed to a proportion (Cribari-Neto & Zeileis, 2010; Ferrari & Cribari-Neto, 2004). However, given that both the compliance rate and absorption rate variables include observations with a 1 (i.e., 100%) value, for these two models, it is necessary to implement fractional regression.

Table 1. Descriptive statistics.

Variable	N	Mean	SD	Minimum	Maximum
Compliance rate	173	0.99	0.02	0.85	1
Absorption rate	173	0.65	0.14	0.14	0.95
Achievements rate	171	0.63	0.19	0.04	1
Quality of government index (EQI)	173	0.28	0.82	−2.24	2.78
Regional autonomy index (RAI)	173	12.90	7.23	1	26
Cohesion policy regional autonomy index (CPRAI)	173	2.42	1.17	1	4
EU co-finance rate (log)	173	−0.66	0.34	−1.49	−0.16
ERDF share GDP (log)	173	7.50	1.41	2.93	10.47
ERDF total (log)	173	19.60	1.26	14.96	22.65
GDP change 2007–11 (pps)	173	0.00	0.03	−0.07	0.07
EU-13 member state	173	0.19	0.39	0	1
Infrastructure share	173	12.7	11.7	0	46.9
Compliance audits	173	1.27	1.18	0	4
Crisis flexibility	173	0.60	0.49	0	1

4. EMPIRICAL RESULTS

The results of the regression analysis are illustrated in Table 2. In line with theoretical expectations, the core finding is that the quality of regional government is a significant predictor of Cohesion Policy administrative performance. Importantly, the positive and statistically significant relationship holds for all three dimensions of administrative performance (compliance, absorption and achievements). The greater the quality of government, the higher the compliance rate, the spending rate and the rate of achieved outcome targets. Further, this relationship holds for both the 2010 and 2013 versions of the regional EQI providing further credibility in the robustness of the relationship.

Table 2. Explaining compliance, absorption and achievements.

	Compliance		Absorption		Achievements
Quality of government index (EQI)	0.111**		0.310***		0.176***
	(−0.041)		(−0.073)		(−0.053)
Regional autonomy index (RAI)	−0.019		0.031		−0.021
	(0.002)		(−0.018)		(0.015)
Cohesion policy regional autonomy index (CPRAI)	−0.011		0.006		−0.108
	(0.123)		(0.149)		(0.116)
Interaction: RAI × CPRAI	0.002		−0.008		0.011
	(0.008)		(0.008)		(0.006)
EU co-finance (log)	(0.459)		(0.386)		(0.061)
	(0.330)		(0.262)		(0.107)
ERDF share of GDP (log)	0.037		0.104*		0.067
	(0.068)		(0.049)		(0.060)
ERDF budget (log)	−0.084		0.042		−0.168*
	(−0.095)		(0.057)		(−0.054)
GDP change 2007–11	1.958		8.044*		0.0591
	(1.932)		(3.538)		(2.819)
EU-13 member state	−0.346		0.306		0.381
	(0.322)		(0.257)		(0.245)
Infrastructure share	0.004		−0.001		0.007
	(0.005)		(0.005)		(0.005)
Compliance audit (reservations)			−0.198***		0.070
			(−0.049)		(0.051)
Crisis flexibility			0.557**
			(−0.22)
_constant	4.176*		−1.201		3.035**
	(−1.596)		(−1.090)		(−0.914)
Observations	173		173		171

Note: Robust standard errors (in parentheses) are clustered by country. *p < 0.05, **p < 0.01, ***p < 0.001.

Turning to the control variables, four of the other factors modelled have an impact on administrative performance (relating to EU monitoring, financial interests and the economic crisis) but inconsistently across the three dimensions of compliance, absorption and achievements.

Regional autonomy is not significantly associated with the administrative performance of regional programmes. This finding corroborates and strengthens the findings of Mendez and Bachtler (2017) with respect to financial compliance and Incaltarau et al. (2019) on absorption, extending the conclusion to performance management of outcomes for the first time. Moreover, the impact of the more policy-specific measure of Cohesion Policy regional autonomy on administrative performance is also statistically insignificant. Regional programmes with a greater level of regionalization of management authority did not perform better than programmes with centralized management arrangements. We also tested an interaction effect to see whether the effect of political regional autonomy on administrative performance was conditioned by Cohesion Policy regional autonomy, again yielding insignificant results.

Goodness-of-fit does not have a significant effect on compliance, spending performance or the achievement of targets. This contrasts with the findings of Mendez and Bachtler (2017) who found a positive link between goodness-of-fit (using a different proxy relating to the share of EU funds in public investment) and the level of compliance (based on the rate of financial corrections) at the national level (instead of the regional level).

By contrast, EU monitoring of compliance is associated with absorption but not with financial compliance or achievement of targets. The significant and positive relationship between EU monitoring and absorption may seem unsurprising given that reservations/interruptions placed on programmes temporarily block EU payments until the underlying issues are resolved. However, it also suggests there is a trade-off between the demands of more onerous audit and oversight of compliance on the one hand and the need to spend funding on the other.

EU financial interests (i.e., the scale of regional ERDF allocations relative to the overall EU ERDF budget) are closely related to EU monitoring, given the greater potential for the largest EU programmes to contribute to overall EU goals or, conversely, to pose reputational risks from poor performance. The results show that the EU financial interests variable is not significantly associated with compliance or spending and, surprisingly, is negatively associated with the achievement of outcome targets: the higher the funding allocation to regional programmes, the lower the target achievement rate. A potential explanation for this counterintuitive finding meriting further investigation could be that the greater pressure on larger programmes to achieve results leads to regional programme managers setting overly ambitious targets during the planning stage that are subsequently difficult to achieve in practice.

The economic and financial crisis of 2008–10 impacted on spending performance, with lower (higher) absorption rates found in regions with slower (faster) rates of economic growth in terms of GDP per capita (pps). These findings are intuitive but are inconsistent with the findings of Tosun (2014). However, the impact of the crisis on growth is not associated with either compliance or the achievement of outcome targets. Related to this, administrative flexibilities introduced by the EU to facilitate absorption during the crisis – by reducing the need to provide domestic co-financing temporarily – increased spending rates in those regional programmes that were eligible for this crisis flexibility measure, as reflected in the positive association between co-financing flexibility and the spending rate.

Lastly, we did not find a statistically significant relationship between the remaining two control variables (EU membership experience and infrastructure spending) and administrative performance. Regional programmes in the EU-13 member states (with less experience in administering EU funding) and regional programmes with higher levels of infrastructure spending (typically involving larger scale and more complex procedures and investments) did not correlate with financial compliance, absorption or achievement of targets.

Overall, the results of the regression analysis confirm our core theoretical hypothesis. The quality of regional government is a significant predictor of Cohesion Policy administrative performance in terms of financial compliance, absorption and achieving outcome targets. An illustration of the relationship between quality of government and each of the three measures of administrative performance is provided in Figures 2–4, while controlling for other variables in the models. A visual inspection of the three figures confirms – in line with the regression results – that absorption performance has a stronger relationship with quality of government than financial compliance or achievement of targets.

Figure 2. Effect of the quality of government on compliance.

Figure 3. Effect of the quality of government on absorption.

Figure 4. Effect of the quality of government on outcome achievement.

5. CONCLUSIONS

This article provides the first analysis of the relationship between the quality of government and multiple dimensions of administrative performance in Cohesion Policy across EU regions. Regression analysis was undertaken to estimate the impact of the European QGI on compliance, absorption and achievement of targets in 173 regional programmes, while controlling for a range of other factors (institutional and policy-specific regional autonomy, goodness-of-fit, EU financial interests/monitoring, funding scale, funding interruptions, economic crisis, EU membership experience, type of expenditure).

There have been several studies that demonstrate a relationship between quality of government and absorption of Structural and Investment (Cohesion) Funds at national level and in some case study regions (e.g., Baun & Marek, 2017; Milio, 2017; Terracciano & Graziano, 2016). Our findings show that absorption performance is also related to quality of governance at regional level across the EU. Further, the research also shows a relationship between quality of government and two other measures of implementation performance – regulatory compliance and achievement of outcome targets. The control variables used in the analysis allow several further conclusions to be drawn on implementation and performance.

First, regional autonomy does not influence regional administrative performance for EU funds. Although it is well-established that territorial governance arrangements can influence implementation outcomes, there is no evidence in our research of regional autonomy influencing administrative performance under any of the three measures. Moreover, this finding applies both to the political RAI and a more targeted Cohesion Policy RAI measuring the level of regionalization of programme management responsibilities.

A likely explanation is the intensified use of EU conditionalities to influence performance over successive programme periods, applied to absorption (the decommitment rule), compliance (the audit ‘explosion’), and – to a lesser extent – achievements (performance reserve) for both national and regional programmes. In other words, the scope for variability across regional programmes with different territorial governance has decreased (Bachtler & Ferry, 2015; Bachtler & Mendez, 2020b; Mendez & Bachtler, 2014). A further factor may be the increasing centralization and national influence over Cohesion Policy spending, again limiting the scope for regional variation (Bachtler & Mendez, 2020b).

Second, EU compliance monitoring was too strongly focused on financial absorption in 2007–13 at the expense of the effectiveness of spending. The positive relationship between EU compliance monitoring and absorption contrasts with the negative relationship for achievements. This is significant given the findings of previous research that less administrative capacity is required for managing the absorption of funding than ensuring it is effective. Aivazidou et al. (2020, p. 1) found that low administrative capacity of regional strategies ‘may increase the absorption rate, though without supporting regional growth’. In this respect, our findings provide quantitative support for the qualitative research conducted for the 2007–13 period which found programme management dominated by a focus on financial absorption at the expense of outcomes (KMPG & PROGNOS, 2016; Mendez, 2011).

As noted above, the results imply a trade-off between the administrative demands on programme authorities in ensuring spending versus financial compliance. We found that regional programmes subject to ‘reservations’ for non-compliance (involving suspensions of funding and requiring corrective actions to permit payments) had lower spending rates. This provides new evidence of the Cohesion Policy audit explosion thesis (Mendez & Bachtler, 2011, 2014), which highlights the negative and often unintended consequences from the rise in EU Cohesion Policy audit activity since the mid-2000s. The finding also indicate that EU compliance monitoring gave insufficient attention to how well EU funds are used and justify the increasing efforts made by the European Commission to strengthen the performance framework.

This differential importance accorded to the three indicators of administrative performance may also explain the impact of the economic and financial crisis (see also Incaltarau et al., 2019; Hagemann, 2019). When the crisis hit, the priority of the European Commission was to maintain spending rates, at a time of falling demand for investment support especially from the private sector. Packages of measures to simplify programme administration were introduced and were largely successful in accelerating spending (Bachtler & Mendez, 2010; European Council, 2010; Mendez & Kah, 2009). Indeed, the statistical analysis of regional programmes confirmed that spending performance was greater among those programmes that introduced increased EU co-financing rates as part of the crisis response. However, relatively few programme authorities took advantage of Commission measures to simplify audit procedures or performance indicators, primarily because of the complexity of the programme management systems put in place after extensive negotiations with the Commission when programmes were launched in 2007–08 (Bachtler & Mendez, 2010).

Third, there is no evidence of a performance deficit among the EU-13 member states. This finding differs from the results of other studies which found that EU membership experience is associated with financial absorption, albeit with contradictory findings in terms of the direction of the relationship (Incaltarau et al., 2019; Tosun, 2014), and with compliance (Mendez & Bachtler, 2014). As a group, the EU-13 had better absorption performance than the EU-15 in 2007–13 (KPMG & PROGNOS, 2016), but there were no clear patterns at national level: financial absorption rates were among the highest in the EU in Poland, Estonia, Latvia, Lithuania and among the lowest in Croatia, Romania and Malta. This lack of a relationship is now evident at regional level also. It reinforces the findings of research on the 2004–08 period that for the (then) new member states ‘administrative capacity for implementing Cohesion policy developed faster and better than policy-makers and academic commentators predicted’ (Bachtler et al., 2014, p. 752; see also Sedelmeier, 2012). This conclusion is, however, for the EU-13 as a whole, and there are other indicators that present a more critical picture of EU-13 administrative capacity in areas such as public procurement, especially in individual countries (European Commission, 2017c, 2022).

Fourth, an interesting finding from the analysis is that there is a positive relationship between scale of EU funding and absorption. It is unsurprising that there is pressure on programmes that account for a relatively larger share of the economy both from national government authorities and the European Commission for reputational reasons. More surprisingly, the total value of funding in absolute terms was not associated with stronger compliance or absorption performance; and it was even negatively associated with outcome achievements. A possible explanation for the latter is that pressure on larger programmes to achieve results leads to regional programme managers setting overly ambitious targets during the planning stage that are subsequently difficult to achieve in practice (especially in a period which included the economic and financial crises).

Lastly, our findings reinforce the need for regional development policies to build subnational administrative capacity (European Commission, 2017; OECD, 2018) especially in the light of concern that ‘existing EU-funded tools are only marginally addressing capacity-building of relevance to LRAs [local and regional authorities]’ (European Committee of the Regions, 2018, p. 65). They also respond to the pleas for better understanding of influences on subnational capacity for regional development, including how to ‘strike a balance between performance, compliance and administrative costs’ (OECD, 2018, p. 9).

Given the weaknesses in EU compliance monitoring, and the questionable achievements of larger regional programmes, there is clear evidence to justify the enhanced performance framework introduced by the European Commission for the 2014–20 period, notably the emphasis on more rigorous target-setting. A key area for further research will be to compare the results from this study with the administrative data emerging from implementation of the 2014–20 period to assess whether the greater attention given to outcomes during the planning process does lead to increased effectiveness of programme spending among regional programmes. Finally, while our main goal has been to test the administrative capacity hypothesis on multiple dimensions of performance for EU regional programmes, future research should incorporate regional ‘political’ explanations combing regional- and national-level data on elections, political regionalism and executive–legislative relations.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge support for the research conducted for this paper from the European Investment Bank University Research Scholarship (EIBURS), ‘Administrative Capacity-Building and EU Cohesion Policy’, as well as advice from Hugh Goldsmith, Pascal Boijmans and Denisa Perrin during the EIBURS project. The authors are also grateful for the research assistance provided by Anke Gratz and Jayne Wilson, and acknowledge the helpful comments from three anonymous reviewers.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. These are targets for the immediate outcomes of funding specified in programme documents, and monitored at programme and EU levels. These are the short-term results of programme support and distinct from the ultimate medium/long-term impacts of the funds.

2. An irregularity is an act that does not comply with EU rules and which has a potentially negative impact on EU financial interests, but which may be the result of genuine errors committed either by beneficiaries claiming funds or by the authorities responsible for making payments.

3. Financial corrections involve the deduction of irregular expenditure from the member state payment claim and through withdrawal or recovery of funding from beneficiaries following the detection of irregular expenditure by the member state or the Commission. Mendez and Bachtler (2017) used the financial correction rate as a proxy for non-compliance. However, this measure is not available at the regional programme level. In any case, the net error rate used in this study is the main measure used by EU institutions to quantify the financial risk to the EU budget arising from non-compliant expenditure. The error rate is an audit authority’s estimate of the part of the expenditure for each programme (or group of programmes) which is not legal and regular and is validated by the Commission. The rate is established on the basis of a statistical sampling approach that is representative for the expenditure incurred for the programme (or group of programmes).

4. DG Regio ‘data for researchers’, see https://ec.europa.eu/regional_policy/en/policy/evaluations/data-for-research/.