Chapter 4. Macroeconomic frameworks and institutional factors for regional economic performance

Beyond regional policies, territorial economic outcomes are also affected by factors that are beyond the control of regional actors. These include national or global macroeconomic trends, as well as certain national policies and frameworks that may have a specifically important spatial dimension. This chapter first explores how macroeconomic frameworks and national structural policies have contributed to differences in regional economic performance. It thereafter explores more generally how institutional and governance settings affect regional economic outcomes.


Chapter synopsis

At the national level the 2007-08 crisis has led to a renewed focus on sound macroeconomic framework conditions and the role of structural reforms in ensuring that economies are competitive and resilient to adverse shocks. At the same time there is much less attention in the debate on the role that national policy frameworks play in reducing or reinforcing interregional disparities.

Within countries or single-currency areas such as the euro area, wage growth that is disconnected from productivity growth can lead to imbalances in regional competitiveness. If regional wages grow faster than productivity, unit labour costs rise and the competitiveness of the tradable sector is reduced. In contrast, regions where wages grow slower than labour productivity enhance their competitiveness, which can negatively affect other regions within the fixed exchange-rate area. This effect can be indirect, i.e. rising wages in the non-tradable sector that are not supported by productivity growth lead to higher prices for non-tradable services both for consumers and (tradable) firms.

The analysis of economic performance at the regional levels shows that a 1 percentage point increase in the growth rate of unit labour costs is, on average, associated with a 0.3 percentage point decrease in the growth rate of value added per capita and 0.4 percentage point decrease in exports per capita. One of the reasons why unit labour costs can differ across regions of the same country is centralised wage-setting mechanisms, in combination with persistent differences in productivity growth. For example, wage bargaining at the national level can lead to wage levels that are too high for regions with low productivity growth and too low for regions with high productivity growth. Sufficient flexibility in centralised wage bargaining, including opt-out clauses for struggling firms, can help to bring wages in line with productivity levels.

Flexible regulations that account for the needs of workers, the unemployed and firms in different regions are particularly important for productivity growth in lagging regions. European regions with the lowest per capita gross domestic product (GDP) and those with low per capita GDP growth benefitted more than other European regions from reforms to employment protection of regular contracts in terms of productivity growth over the 2000-13 period. Similarly, product market regulations in wholesale and retail trade appear to have particularly negative impacts on the productivity growth of a country’s least productive regions. Structural reforms should be undertaken preferably during periods of high economic growth. However, as many recent reforms occurred during a period of severe economic weakness, these reforms created higher social costs in terms of job losses than they would have if they had been implemented during a boom period.

Governance and the efficient functioning of public administration can contribute to narrowing productivity gaps. A barrier to the well-functioning of cities is a fragmented administrative structure. A doubling of the number of municipalities within the boundaries of a metropolitan area is associated with 3-6% lower productivity of its workers. This penalty is alleviated by about half when metropolitan governance arrangements exist. The adverse effect of fragmentation is not limited to the city itself, as economic growth in highly urbanised regions with a larger number of municipalities per capita is lower than in less fragmented regions. Judicial efficiency can also play a role. Evidence shows that employment and turnover in Italian firms and thee likelihood of participating in global value chains are lower for firms located in jurisdictions where legal proceedings take longer than in firms located in more efficient jurisdictions.

Macroeconomic framework and structural reforms: accounting for regional differences

Regions are affected by a range of nationwide economic policies and national or even global macroeconomic trends. Policies that are designed to ensure equal treatment across space, e.g. rules that govern the dismissal of workers or the licensing of new firms, can have dissimilar economic effects in different regions. Economic structure, levels of economic development, characteristics of the local labour force and the availability of natural resources are all among the factors that may contribute to the unequal impact of policies on different regions. Macroeconomic developments such as exchange rates and nominal interest rates are – by definition – also out of the control of the individual firm or even region. Regional price competitiveness will hence typically fluctuate – at least in the short run – with changes in nominal exchange rates and relative levels of inflation in the country (or monetary zone, i.e. euro area) to which it belongs. Similarly, if inflation in a region is different from inflation rates in other parts of its country (or its monetary zone), this will affect regional price competitiveness. Depending on their economic structure, the reliance of regions on pure price competitiveness differs greatly – and effects from the aforementioned macroeconomic developments will hence affect them to different degrees.

Regional variations in the impact of policies and macroeconomic trends are amplified by the disinclination of many people to move. Even though there are no legal barriers for people to move to another region in their country, doing so typically has social costs (at least initially), and also requires solving many practical issues. The result is that people tend to prefer to stay in the region where they live. In a recent study in the United States, for example, 89% of applications sent through an online job portal were found to have been sent to firms in the state where the applicant resided (Marinescu and Rathelot, 2016[100]). Estimates for other countries, e.g. for the United Kingdom, suggest that this “distaste” for distant jobs may be even stronger outside of the United States (Manning and Petrongolo, 2017[101]). From a well-being perspective, it is easy to understand why the vast majority of people choose to stay as close to home as possible when looking for work. It implies, however, that people are not necessarily moving to the places where they could be the most economically productive, and also that forces that would tend to equilibrate a potentially uneven impact of national policies or macroeconomic developments among regions can at best be partially effective.

Misalignment between wage and productivity growth - Unit labour costs as a measure of imbalances

Chapter 2 described how regions with a larger tradable sector have been more successful in achieving sustainable economic and productivity growth. However, tradable sectors – by definition – are also directly exposed to international developments – in particular changes in exchange rates and trade shocks. This can be a particular challenge for regions that are highly dependent on production in “mature” sectors, i.e. those producing standardised goods for which technological replication is relatively easy, e.g. textile manufacturing. Cost competitiveness is often essential for such sectors and for the economic vitality of the regions where they are located. Without innovation and the introduction of new products or processes, firms need to rely on keeping unit costs in line with international competitors to remain profitable.

A good example of this is Portugal’s Norte region. The appreciation of the euro, which occurred at the same time as The People’s Republic of China (“China” hereafter) joined the World Trade Organization, jolted the textile manufacturing sector. As a consequence, Portugal’s Norte region was seriously affected; China’s textile exports were directly competing with many of the producers in Norte, and the appreciation of the euro meant that exports from Norte were becoming relatively more costly. The result of the loss in its competitive position was a slow decline in manufacturing employment in the Norte region (c.f. “Is a large tradable sector more risky for a region?”).

Among the different factors that affect production costs, a main focus in the public debate often centres on the cost of labour, including workers’ basic wages or employers’ additional social contributions. While this obviously is a key factor in industries that mainly compete on price, other factors also contribute to a firm’s production costs. These include purchasing and maintenance of machinery, rent for office space, electricity and other utility costs, the time it needs to deal with administrative requests, interest rates at which the firm can borrow, and the cost of imported intermediate products.

A measure that captures the development of personnel costs in relation to productivity is “unit labour costs”. It is calculated as the ratio between personnel costs per employee and labour productivity (i.e. real output per worker). Unit labour costs increase if the compensation of employees rises faster than their productivity and decrease if productivity gains outpace compensation. In the popular debate, the focus is often on rising labour costs. What unit labour costs highlight is that rising labour costs are not necessarily a concern, quite the contrary as they reflect rising incomes and living standards. But if rising labour costs are not offset by productivity gains, increases become easily unsustainable. Unit labour costs can therefore serve as a possible warning sign for accruing imbalances in an economy.

Unit labour costs growth before the 2007-08 crisis

The European countries hit hardest by the 2007-08 crisis were the ones where unit labour costs had risen steadily and in many cases rapidly in the run-up to the crisis (c.f. “The global 2007-08 crisis uncovered some unsustainable growth models” in Chapter 1). While some growth in unit labour costs is natural and driven by inflation, excessive growth is red flag that imbalances are accruing. Annual increases in worker remuneration typically compensate for inflation. But in many countries, personnel costs appear disconnected from productivity growth and inflation. For example, the 3-year growth in unit labour costs in Greece, Spain or Romania exceeded the 9% threshold (12% for non-euro area countries) set by the European Macroeconomic Imbalance Procedure (MIP) Scoreboard in each year between 2003 and 2007.

Over the same time period, the opposite process took place in Germany and other northern European countries. From 2000 to 2007, labour productivity in Germany grew by 11% whereas real wages increased by only 1%. This widened the gap in labour unit costs across European regions and worsened the competitive position of regions with high unit labour cost growth.

Recent empirical OECD work confirms that increasing unit labour costs can be detrimental to regional growth. Over the period 2000-13, regions’ growth in income and export per capita is negatively associated with increases in unit labour costs. A 1 percentage point increase in the growth rate of unit labour costs is, on average, associated with a 0.3 percentage point loss in growth of gross value added per capita and a 0.4 percentage point loss in exports per capita.1 This effect can also be seen for regions in individual countries. For example in Portugal (excluding Lisbon)2 per capita exports in each region increased faster than the country average in years where its unit labour costs were declining, but saw a relative export decline in years when its unit labour costs were rising (Figure 4.1).

Figure 4.1. Rising unit labour costs and falling exports in Portuguese regions (excluding Lisbon), 2000-10
Annual average export growth per capita vs annual average growth in unit labour costs

Note: Relative growth in per capita exports is the difference between the unweighted average of growth rates across regions within the group and the country average growth rate. Growth rates are classified as “strong decline” or “strong growth” are those deviating by more than 1 percentage point from the country average growth rate in unit labour costs, growth rates within 1 percentage point are classified as “decline” or “growth”.

Source: Lembcke, Oliveira Martins and Wolf (forthcoming) based on Eurostat Regional Statistics [Database] and data provided by Los and Chen (2016[76]).

Wages grew faster than productivity, eroding unit labour costs in southern European regions

Changes in unit labour costs in southern European regions prior to the crisis were driven by wages that grew faster than productivity. In this respect, the distinction between tradable and non-tradable sectors turns out to be essential in understanding both the drivers of increases in unit labour costs and the challenge regions face in supporting tradable activities. In regions in Greece, Italy, Spain and Portugal, it was the unit labour costs in non-tradable sectors that drove aggregate unit labour costs.3 In these countries, unit labour costs tended to be not only high in non-tradable sectors, but also tended to grow faster than in tradable sectors.

Strong growth in the costs for staff that are not supported by productivity growth, tend to be particularly damaging in countries that have a fixed exchange rate regime. In a fixed exchange rate regime, such as the euro area, rising unit labour costs cannot be compensated by a depreciation of the nominal exchange rate. Instead, they directly translate into changes in the real effective exchange rate, i.e. a decrease in the country’s cost competitiveness.

Labour costs in non-tradable and tradable sectors tend to move in tandem

Considering separately the two constituent elements of unit labour costs, labour costs and labour productivity, sheds further light on the origins of low regional performance. First, while productivity in tradable sectors in southern and Eastern Europe had been growing before the 2007-08 crisis, productivity in non-tradable sectors had been stagnant, if not declining. These patterns should have been reflected in, at best, moderate labour costs increases, and if anything higher increases in the tradable sector (as workers can move between the two sectors, wages tend to increase in tandem in both sectors). This was indeed the case in most European regions where labour costs for tradables and non-tradables increased at similar rates by and large (see Figure 4.2, left panel). However, the reality was very different in southern European regions with low growth where labour costs in the non-tradable sector increased in line with the rest of the country, and in particular much faster than labour costs in the tradable sector (Figure 4.2, right panel). In such a situation, workers tend to move from the tradable to the non-tradable sector, resulting in a decline in the tradable sector while the non-tradable sector increases strongly – usually in an unsustainable fashion.

Figure 4.2. Labour costs in tradables and non-tradables strongly diverged in some regions
Relative labour costs in low growth and low income regions compared to other regions in the country separated for tradable and non-tradable sectors (100=non-tradable sectors in 2000 in regions that are not low income or low growth), 2000-13

Note: Personnel costs are the total compensation of employees from regional accounts.

Source: Lembcke, Oliveira Martins and Wolf (forthcoming) based on Eurostat Regional Statistics [Database].

Labour cost misalignments can be driven by capital flows, but labour market institutions may also play a role

Normally, market forces contribute to the alignment of non-tradable and tradable labour costs in a given labour market. It is hence somewhat surprising to see stark divergences in labour costs for the non-tradable and tradable sector within the same region. To understand such a situation, it first needs to be acknowledged that prices in the tradable sector need to take international prices for the same goods and services into account, which does not give companies much flexibility in wage setting. International competition implies that if wages increase too much (i.e. significantly above increases in productivity), such companies will go out of business. As a result of this competition, tradable firms’ wages should reflect the productivity of their employees.

In contrast, firms in non-tradable sectors are (at least in the short- to medium-term) unconstrained by international or global prices since these firms do not face competition from competitors in other regions. In situations of strong demand for non-tradables – such as when demand is driven by strong capital inflows into a country or region – prices and wages in the non-tradable sector may hence see significant increases, even if those often turn out to be unsustainable in the medium- to long-term. Such increases may then become self-perpetuating, as in a situation of comparatively higher inflation in a given region (or country within a monetary union), real interest rates may become negative and – through fuelled credit demand – further inflate an already booming non-tradable economy. Such a boom can go on for some time until eventually capital flows into the region or country dry up and a tightening credit supply ends the party.

Furthermore, if wages are sectorally negotiated for a whole country (as is the case in many European countries) and do not account for local cost of living differences, the real wage in non-tradable sectors – including the public sector – can easily become too high in low cost regions and too low in high cost regions.4 The same is not necessarily true for tradable sectors. The reason is that every place needs hairdressers, kindergarten teachers or construction workers, but the production of textiles, cars or development of computer programmes is not spatially bound. Thus, whereas economic activity in the tradable sector will move to other regions in the long run if unit labour costs are too high, firms in the non-tradable sector will stay in the region.

The difference in wage setting regimes between the south and the east of Europe is a potential explanation for the differing growth performance of southern and eastern European regions. Across countries and industries there is evidence that the misalignment of wage and productivity developments is greater where wages are not set at the firm level (ECB, 2015[102]). In contrast, centralised bargaining might have dampened the adverse employment effects following the 2007-08 crisis. There is some evidence that employment in countries with stronger centralisation of collective bargaining fell less as wages in these countries adjusted quicker to the shock (OECD, 2017[96]).

Flexible institutions and structural reforms might be particularly relevant in lagging regions

Flexible labour market regulations that account for the needs of workers, the unemployed and firms in different regions are particularly important for lagging regions. Increased flexibility in wage setting and reforms of Germany’s social security and welfare system have been credited with the expansion of employment and a decline in unemployment since 2003, dubbed the “German labour market miracle” (Burda, 2016[103]). Estimates show that rigid employment regulations can hurt productivity growth more in lagging regions than in regions that are already more productive and less able to cope with more rigid labour market regulations (D'Costa, Garcilazo and Oliveira Martins, 2016[104]).

Low-income and low-growth regions experienced stronger productivity growth than other European regions following labour market reforms (Figure 4.3). However, major labour market reforms are rare and changes often piecemeal or targeted at fixed-term contracts. The OECD strictness index for employment protection legislation (EPL) in Italy, for example, did not change for more than 20 years until the contentious “Monti-Fornero reform” was introduced in June 2012. Despite the political effort involved in passing the reform, only three out of the 25 constituent indicators considered in the OECD Indicators of Employment Protection changed.

Figure 4.3. Flexibility in the labour market can boost productivity growth
Cumulative impact of a decrease in the stringency of Employment Protection Legislation on productivity growth

Note: Estimates based on multivariate regression allowing for a delayed impact of reforms to the labour market up to two years.

Source: D’Costa, Lembcke and Oliveira Martins (forthcoming).

Similarly, product market regulations in wholesale and retail trade appear to have particularly negative impacts on the productivity growth of a country’s least productive regions (those farthest from the leading region of the country in terms of GDP per worker), which includes Europe’s lagging regions. Conversely, trade openness appears to help less productive regions disproportionately more than other regions, particularly in low-growth countries (D’Costa, Garcilazo and Oliveira Martins, 2013[105]). It is therefore important to consider not only the aggregate impact of nationwide structural policies, but their impact on different types of regions and particularly on lagging regions.

Institutions and governance: Constraints and catalysts

Beyond national macroeconomic trends or structural reforms, regional growth and productivity is also more generally affected by a country’s institutional and governance settings. These settings can be a facilitator or barrier to regional economic performance. Institutions and, more generally, social norms provide the foundation for the interactions of firms, workers and consumers among each other and with the government. Of particular importance are institutions that guide these economic interactions, such as the rule of law and low levels of corruption.5

While much of the evidence on the effect of institutions is focused on developing countries, differences in the quality of government across European regions – and in particular low levels of corruption and a high degree of accountability – seem to be highly relevant in explaining per capita GDP growth.6 The European Quality of Government Survey provides several waves of data on good governance in the public sector and illustrates that the difference in governmental quality can be stark, not only across but also within countries.7 For example, the best-performing Spanish and Italian regions rank among Europe’s highest scoring regions, but other regions in the countries fall below the EU average. Low quality of government in a region is highly correlated with the level of economic development in the region, but also with levels of social trust (Charron, Dijkstra and Lapuente, 2014[106]).

Fragmented governance curtails productivity of cities and urban regions

Fragmented governance arrangements can curb productivity and economic growth in cities and regions. Ahrend et al. (2017[107]) estimated that a doubling of the number of municipalities within the boundaries of a metropolitan area was associated with 3-6% lower productivity, but that this penalty is alleviated by about half when metropolitan governance arrangements exist. The adverse effect of fragmentation is not limited to the city itself, but affects the whole region. Economic growth in highly urbanised regions with more municipalities per capita is lower than in regions with less administrative fragmentation; for regions with a mostly rural population there is, however, no statistically significant impact of increased administrative fragmentation (Bartolini, 2015[108]). 8

The impact of administrative fragmentation on productivity most likely arises through inefficiency in policies that require co-ordination across different local or regional governments, examples being transport and land-use planning, which also should be aligned with tax and environmental policies (OECD, 2017[109]) or growth-promoting policies in general (Cheshire and Magrini, 2009[110]). For example, administrative fragmentation in the Chicago Metropolitan Area has been one of the factors behind an overly complex and not particularly efficient governance structure of public transit providers in the metropolitan agglomeration. This, in turn, has been reflected in relatively low levels of integration of the public transit system, and has also contributed to underinvestment in its infrastructure (OECD, 2015[111]).

Improving administrative efficiency has significant economic returns

Barriers to productivity growth and trade integration can extend to judicial efficiency. The length of legal proceedings can vary significantly between jurisdictions. In Italy, the average length of civil proceedings in southern regions was twice that of courts in the north for cases in the 2002-07 period.9 But even in Italian cities like Milan or Rome, the average time to reach a ruling is more than double that in Madrid (Spain) or that in Polish regions.10 But even within the north and the south, judicial efficiency varies significantly. Employment and turnover of manufacturing firms has been estimated to increase by 2% for a 10% reduction in the length of court proceedings (Giacomelli and Menon, 2017[112]). Firms in less efficient jurisdictions are also less likely to participate in global value chains. The probability of a manufacturing firm supplying specialised inputs abroad increases by about 0.4 percentage points for a 10% reduction in average trial length (Accetturo, Linarello and Petrella, 2017[113]).

The ease of doing business varies not only across, but also within countries

Judicial aspects are not the only factor that varies across regions. The World Bank’s “Ease of Doing Business” indicators show that national frameworks and general business conditions can limit business dynamics, with obvious negative effects on growth and productivity. For example, in countries with low growth regions, the most recent range of indicators stresses how firms struggle to access credit, enforce contracts and to deal with issues involving minority investor protection. In Europe’s eastern countries with low-income regions, insolvency regimes and contract enforcement are key challenges (World Bank, 2016[114]). For several countries, data on the business environment at the subnational level have recently become available, confirming the existence of major differences not only across countries, but also across regions within countries.11


[16] Accetturo, A., A. Linarello and A. Petrella (2017), “Legal enforcement and Global Value Chains: Micro-evidence from Italian manufacturing firms”, Occasional Papers (Questioni di economia e finanza), No. 397, Banca D'Italia.

[31] Acemoglu, D., S. Johnson and J. Robinson (2001), “The Colonial Origins of Comparative Development: An Empirical Investigation”, The American Economic Review, Vol. 91/5, pp. 1369-1401.

[30] Acemoglu, D. and J. Robinson (2012), Why Nations Fail: The Origins of Power, Prosperity and Poverty, Profile Books, London.

[10] Ahrend, R. et al. (2017), “What Makes Cities More Productive? Evidence From Five OECD Countries on the Role of Urban Governance”, Journal of Regional Science, Vol. 57/3, pp. 385-410.

[29] Auricchio, M. et al. (2017), “The consequences of public employment: evidence from Italian municipalities”, Working Papers, No. 1125, Banca d'Italia.

[11] Bartolini, D. (2015), “Municipal Fragmentation and Economic Performance of OECD TL2 Regions”, OECD Regional Development Working Papers, No. 2015/2, OECD Publishing, Paris,

[6] Burda, M. (2016), “The German Labor Market Miracle, 2003 -2015: An Assessment”, SFB 649 Discussion Paper, No. 2016-005, Humboldt -Universität zu Berlin.

[28] Cardoso, A. and P. Portugal (2005), “Contractual Wages and the Wage Cushion under Different Bargaining Settings”, Journal of Labor Economics, Vol. 23/4, pp. 875-902.

[9] Charron, N., L. Dijkstra and V. Lapuente (2014), “Regional Governance Matters: Quality of Government within European Union Member States”, Regional Studies, Vol. 48/1, pp. 68-90.

[27] Charron, N., L. Dijkstra and V. Lapuente (2015), “Mapping the Regional Divide in Europe: A Measure for Assessing Quality of Government in 206 European Regions”, Social Indicators Research, Vol. 122/2, pp. 315-346.

[13] Cheshire, P. and S. Magrini (2009), “Urban Growth Drivers in a Europe of Sticky People and Implicit Boundaries”, Journal of Economic Geography, Vol. 9/1, pp. 85-115.

[8] D’Costa, S., E. Garcilazo and J. Oliveira Martins (2013), “The Impact of Structural and Macroeconomic Factors on Regional Growth”, OECD Regional Development Working Papers, No. 2013/11, OECD Publishing, Paris,

[7] D'Costa, S., E. Garcilazo and J. Oliveira Martins (2016), “Impact of Structural Reforms on Regional Growth: Distance to the Frontier Matters”, SERC Discussion Paper, No. 203, Spatial Economics Research Centre.

[4] ECB (2015), “Economic Bulletin”, No. 08/2015, European Central Bank.

[26] European Commission (2017), “Competitiveness in Low-Income and Low-Growth Regions: The Lagging Regions Report”, European Commission Staff Working Document, European Commission.

[25] Garnero, A. (2017), “The Dog That Barks Doesn't Bite: Coverage and Compliance of Sectoral Minimum Wages in Italy”, Discussion Paper, No. 10511, IZA.

[24] Gaulier, G. and V. Vicard (2013), “The signatures of euro-area imbalances: Export performance and the composition of ULC growth”, CompNet Policy Brief, No. 02/2013, European Central Bank.

[15] Giacomelli, S. and C. Menon (2017), “Does weak contract enforcement affect firm size? Evidence from the neighbour’s court”, Journal of Economic Geography, Vol. 17/6, pp. 1251–1282.

[23] Ketterer, T. and A. Rodríguez-Pose (2016), “Institutions vs. ‘first-nature’ geography: What drives economic growth in Europe's regions?”, Papers in Regional Science.

[3] Los, B. and W. Chen (2016), “Global Value Chain Participation Indicators for European Regions”, Report for the OECD, No. December 2016.

[2] Manning, A. and B. Petrongolo (2017), “How Local Are Labor Markets? Evidence from a Spatial Job Search Model”, American Economic Review, Vol. 107/10, pp. 2877-2907.

[1] Marinescu, I. and R. Rathelot (2016), “Mismatch Unemployment and the Geography of Job Search”, NBER Working Papers, No. 22672, National Bureau of Economic Research, Cambridge, MA.

[22] Neumark, D. and W. Wascher (2011), “Does a Higher Minimum Wage Enhance the Effectiveness of the Earned Income Tax Credit?”, Industrial and Labor Relations Review, Vol. 64/4.

[21] Neumark, D., M. Thompson and L. Koyle (2012), “The effects of living wage laws on low-wage workers and low-income families: What do we know now?”, IZA Journal of Labor Policy, Vol. 1/1, pp. 1-34.

[14] OECD (2015), Governing the City, OECD Publishing, Paris,

[32] OECD (2016), OECD Regional Outlook 2016: Productive Regions for Inclusive Societies, OECD Publishing, Paris,

[12] OECD (2017), The Governance of Land Use in OECD Countries: Policy Analysis and Recommendations, OECD Publishing, Paris,

[5] OECD (2017), OECD Employment Outlook 2017, OECD Publishing, Paris,

[20] World Bank (2013), Doing Business in Italy 2013: Smarter Regulations for Small and Medium-Size Enterprises, World Bank Group, Washington, DC.

[18] World Bank (2015), Doing Business in Spain 2015, World Bank, Washington, DC.

[19] World Bank (2015), Doing Business in Poland 2015, World Bank, Washington, DC.

[17] World Bank (2016), Doing Business 2016: Measuring Regulatory Quality and Efficiency, World Bank, Washington, DC.


← 1. Cited results from Lembcke, Oliveira Martins and Wolf (forthcoming) account in the regression for time-invariant regional characteristics, the change in the skill composition of the local labour force and trends at the country level. Gaulier and Vicard (2013[188]) assess the impact of unit labour cost increases at the country level and find little relationship with exports, but a strongly negative relationship with the current account deficit.

← 2. Excluding the archipelagos.

← 3. The three-year growth rate in non-tradable regional unit labour costs was higher than the national average growth rate in all years between 2000 and the 2007-08 crisis. Greece is the only exception where over the period 2000-03 non-tradable unit labour costs in regions not classified as “low-growth” was below the national average.

← 4. Public sector wages are often also determined centrally and might play a role in explaining the patterns found in this study. Utilising a more stringent definition of “non-tradable” that is dominated by public and related sectors (public administration, defence, health care and social services) shows an even stronger alignment of personnel costs than in the main delineation used here and in the OECD Regional Outlook 2016 (OECD, 2016[1]). The benefits are not necessarily passed on to workers. A decrease in public employment between 2001 and 2011 in Italy increased private sector employment, especially in southern regions (Auricchio et al., 2017[215]). Spatial segregation of low-wage/low-productivity jobs and sectors in some regions and high-productivity/high-cost ones in others can partly circumvent the stringency of centralised wage setting. Another solution is that wage bargaining only constitutes a wage floor, with better-performing firms paying their workers an additional premium. Firms in Portugal have used these “wage cushions” extensively (Cardoso and Portugal, 2005[204]). Firms might also simply not comply with regulations. For instance, Garnero (2017[189]) estimated that about 10% of Italian workers are paid below the minimum wage set by their collective bargaining agreement. This share is higher in Italy’s south where the cost of living is less than in the north where prices are high. The disparity ranges from 8.5% of the workforce being underpaid in the northeast to 18.5% in the country’s south. Instead of aiming to alleviate poverty through general increases in nominal wages (or minimum wages), targeted strategies (such as living wages that account for local cost of living) are likely to be more successful in targeting low-income households and have less adverse effects on employment, as found for the United States (Neumark, Thompson and Koyle, 2012[168]). Direct measures, such as transfers via earned income tax credits that benefit low-income households can directly result in better distributional outcomes for low-income households (Neumark and Wascher, 2011[169]). A further advantage of this type of support is that it ties work incentives to an increase in income without disconnecting employer wages from worker productivity.

← 5. Acemoglu, Johnson and Robinson (2001[224]) made the seminal contribution that empirically linked institutions and economic development. See Acemoglu and Robinson (2012[223]) for a summary of the literature detailing the characteristics of institutions that support growth.

← 6. Low levels of corruption and a high degree of accountability are associated with good performance in terms of economic growth in European regions (Ketterer and Rodríguez-Pose, 2016[178]).

← 7. It surveys residents on their experience and interaction with the public sector in areas such as corruption, accountability (in case of corruption), rule of law and effective service delivery (in education and health care). See Charron, Dijkstra and Lapuente (2015[203]) for a description of the survey.

← 8. Based on data for the 1996-2011 period on per capita GDP growth and administrative fragmentation in 250 TL2 regions from 23 OECD countries. The regression includes education, density and innovation controls, as well as country fixed effects.

← 9. See Giacomelli and Menon (2017[112]) and related sources.

← 10. See the discussion by the European Commission (2017[11]) based on subnational reports on the ease of doing business for Italy (World Bank, 2013[161]), Poland (World Bank, 2015[160]) and Spain (World Bank, 2015[159]).

← 11. Reports are available for Italy (World Bank, 2013[161]), Poland (World Bank, 2015[160]) and Spain (World Bank, 2015[159]). The assessment of the business environment focuses on the main cities in different countries. In aggregate, the three countries with available indicators form a clear ranking. Polish cities tend to have a more conducive business environment, followed by Spanish and Italian regions. With regard to certain indicators, however, there are substantial differences across countries and regions. Starting a business or registering property takes longer in Polish regions than in Italy or Spain, but receiving a construction permit is significantly faster.