Abstract

This paper examines the geographical impact of the recent austerity measures in Greece. Owing to compositional differences across regions, the horizontal measures are found to amplify existing disparities. It is argued that under certain conditions, relating to wider spatial imbalances in the country, this can trigger cumulative divergence processes that may be hard to address in the future. To correct this, it is proposed that revenue-generating efforts should concentrate more on tackling tax evasion and increasing tax progressivity; while the reduction in public consumption should be compensated by targeted increases in public investment. Facilitating the early release of already earmarked European Union funds can be central for such a strategy.

Introduction

Although in the initial stages of the global financial crisis, Greece did not appear particularly affected; by the end of 2009, Greece entered an unprecedented fiscal crisis, which is still threatening the stability of the country and of the European Union's Economic and Monetary Union (EMU) at large. In response to these developments and pushed by its European partners, the Greek government set out to implement an extensive package of austerity measures. As with elsewhere in Europe, these measures are geographically horizontal. Nevertheless, due to regional differences in specializations, incomes and economic capacities, horizontal measures can have significant spatial effects, affecting different regions disproportionately. Indeed, in his analysis of the local geographies of the crisis in the USA and the UK, Martin (2010) argues that public sector cuts “will almost certainly impact some cities and localities much more than others. The ramifications of the financial crisis have proved anything but spatially uniform.” Similarly, Rowthorn (2010) has argued for the UK that that the public sector cuts announced by the coalition government will affect disproportionately the north of the country, for which public sector employment represents a higher proportion of total employment.

This composition issue is particularly important for Greece, due to the country’s multi-faceted inequalities and weak cross-regional equilibration mechanisms (Christopoulos and Tsionas, 2004; Monastiriotis, 2009; Petrakos and Saratsis, 2000). In Greece, more than elsewhere in Europe, economic activity is highly concentrated in a few regions, with Attica, the broader region of the capital Athens, accounting for some 40% of population and almost 50% of national gross domestic product (GDP). Industrial activity is also largely concentrated there as is the incidence of foreign-owned and export-oriented manufacturing (Fotopoulos et al. 2010; Monastiriotis and Jordaan, 2010; Petrakos and Psycharis, 2004). The remaining regions have very low specializations, mainly in tourism (island regions, especially the South Aegean and Crete), agriculture (accounting for over 30% of employment in Thessaly, Peloponnese, Eastern Macedonia and Thrace, Western Greece and parts of Central Greece and Central Macedonia) and light manufacturing (Central Greece and Central Macedonia), with financial and other business services accounting for less than 5% everywhere except in the main urban regions of Athens and Thessaloniki.

Such structural imbalances, and the developmental weaknesses that they manifest, can raise concerns that the austerity measures may have significantly differentiated implications across space—not only in relation to compositional income effects but also with regard to more structural and more permanent effects to the real economy. In countries with strong cross-regional equilibrating mechanisms (migration, capital mobility, price adjustments) and a history of effective policy interventions to address regional imbalances, such a differentiation may not matter in the long run—especially if the austerity shock is transitory. But in a country like Greece and in the context of a prolonged fiscal consolidation programme (already expected to last beyond 2015—Monokroussos, 2011), this differentiation may lead to more permanent divisions across space—perhaps in a cumulative fashion.

Of course, as the implementation of these measures is still unfolding, it is not possible to provide here an accurate measurement of the spatial consequences of these measures and of their long-run implications. Given, however, the lack of attention from the side of policy to the spatial dimension of these issues,1 a preliminary examination of the geographical effects of the austerity measures is particularly important. This is not only in order to provide an early warning to regional policy about the future challenges that it may face but also for evaluating the suitability and effectiveness of the measures at the national level. Although fiscal consolidation is an unquestionable priority in the face of a national default and a possible exit from the Eurozone, it is important that the means for achieving this do not compromise the future economic cohesion of the country by intensifying already pronounced regional imbalances.

With this in mind, in this paper, we pursue two complimentary pieces of analysis. Firstly, we undertake an ex-ante accounting evaluation of the geographic composition of the income effects of the austerity measures implemented in Greece since March 2010. We rely on information from a variety of sources about the distribution of public sector employment, the incidence of low- and high-pay and of tax evasion, and the relative importance for each region of funds distributed through public investment and public transfers. Secondly, we offer an exploratory discussion about how the asymmetric income effects may be translated into longer-run structural imbalances across the Greek regions by elaborating on the relevance and mechanics of a cumulative causation process that can be triggered by these asymmetries. The purpose here is not to predict with any claimed accuracy the regional evolutions of the future, but to identify the possible threats to regional and economic cohesion that the horizontal implementation of the austerity measures may entail. Although this discussion is by its nature specific to the Greek context, some of the issues raised are expected to be of wider relevance to Europe, as the wave of fiscal consolidation measures extends to other countries in the European south and beyond.

The remainder of the paper is structured as follows. The next section examines briefly how the economic crisis spread to Greece and reviews the austerity measures implemented. This is followed by an examination of the direct spatial impact of the measures (compositional effects) and, subsequently, a section exploring the longer run implications of these measures. We conclude with some implications for policy.

The Greek crisis and the austerity programme

As has been discussed extensively in the popular literature, what started in 2007 as a ‘mortgage crisis’ in the USA soon extended to most of the rest of the developed world in the form of a ‘financial crisis’, as uncertainty about who holds ‘toxic assets’ and ‘bad debt’ spread. The ‘liquidity crisis’ that this translated to led to an all-out ‘economic crisis’, with production firms facing increasing difficulties in financing their everyday activities and wider investment plans. In this global context, Greece appeared initially to be well protected. The country had very low exposure to international trade (with goods exports representing a mere 8% of national GDP), a rather vibrant banking system with low exposure to toxic assets and a history of strong growth for over a decade. Participation in EMU seemed at first a blessing, as currency pressures hit mainly countries at the vicinity of the eurozone, while the common currency appeared until the second half of 2009 to provide a safe haven for countries with traditionally weak currencies and fundamentals.

Underneath this, however, two important structural constraints were soon to expose Greece to an unprecedented fiscal crisis. Firstly, Greece’s chronic inability to control its public expenditures and generate sufficient revenues in line with other European countries. Especially on the revenue side, Greece significantly underperformed relative to the European average, with tax revenues as a share of GDP being about 7%-points lower (around 32%) and declining since the early 2000s (Servera and Moschovis, 2008). Weak tax collection mechanisms and pervasive corruption and tax evasion are deep-rooted problems that have systematically contributed to this (Matsaganis and Flevotomou, 2010; Skouras and Christodoulakis, 2011). Secondly, systemic problems in the EMU design which created a structural asymmetry within the Eurozone, resulting in real currency appreciation and continuous loss of competitiveness in the European south.2 Low interest rates, partly due to suppressed wage growth in Germany, led to fast consumption expansion in less competitive countries such as Greece and to asset-price inflation (including a housing bubble). Owing to Greece’s weak industrial base and high product market rigidities, these developments led in turn to accelerating inflation rather than accelerating productivity growth (Mitsopoulos and Pelagidis, 2011). As EMU does not allow for national currency devaluation, this in turn pushed unit labour costs upwards, contributing to a continuously deteriorating current account deficit (which, at 14% of GDP in 2008, had surpassed that experienced by Argentina before its default in 2001) and putting further pressures on the country’s public finances.

Irrespective, however, of these structural imbalances, the crisis was triggered by a more subtle event that had to do with another Greek particularity: the weak monitoring and apparent misreporting of fiscal data. Starting from a forecasted budget deficit of 3.7% of GDP (as reported in the 2009 Convergence Programme in December 2008), successive revisions of the deficit forecasts around the period of the October 2009 elections brought the deficit to 5.4% (October), 10.6% (November) and later 12.7% of GDP (December 2009). The official figure released by Eurostat in November 2010 was a spectacular 15.5% of GDP. In a climate of international financial instability and at least partly owing to the lack of a robust response by the European Union (EU) institutions and member states to the unfolding crisis, this turmoil created first a ‘credibility crisis’ that pushed Greek government bond spreads to unattainable levels (over 1000 basis points in March 2010). This destabilized further the Greek economy, as economic confidence collapsed and fears of a deep recession materialized, putting additional pressures on the government debt and the budget deficit. The situation got out of control by spring 2010 leading to an acute ‘fiscal (sovereign debt) crisis’, with a possible default becoming seemingly inevitable.

Under the fear of the implications that a Greek default, inside the Eurozone, would have politically for the EMU project and financially for the other member states, the EU agreed, together with the IMF and the ECB (the so-called, troika), literally on the 11th hour, an emergency rescue package in the form of a €110bn loan to the Greek government. The loan entailed a set of provisions for the implementation of a range of austerity measures and accompanying structural reforms aiming at recovering public finances and helping the economy regain some of its lost competitiveness. As the public-financial situation worsened and the economy kept sliding into ever-deepening recession, the austerity measures became more pervasive, raising significant public discontent but also weakening further domestic demand and investor confidence. Indeed, concerns about a Greek default continue today, even after the EU agreement to extend the repayment period and reduce the interest rate of the Greek loan and the intensifying efforts since June 2011 to achieve a rollover of the Greek debt with voluntary private sector participation. If anything, this partial debt restructuring has so far been taken by the markets as a signal of increased default risk leading to a further downgrading of Greece by the international credit rating institutions.

It is in this context that the Greek government announced, first in March 2010 and at various stages subsequently, a series of austerity measures aiming at reducing its excessive budget deficit to below the 3% threshold by 2015. The original fiscal consolidation measures of 2010, which aimed at creating savings to the value of 7.4% of GDP, were gradually amended with measures representing a fiscal adjustment equal to €14.4bn (6.3% of GDP) in 2011 and an additional €23bn for the period 2012–2015.3 On the revenue side, the main measures include a rise in VAT (from 19 to 23% for the standard and from 9 to 11% for the basic rate and an expansion to product categories not previously covered) and in taxes on fuel, tobacco and alcohol, an one-off levy of 1% on very profitable firms and high-income households (complemented, more recently, with additional levies to households—most, but not all, of which have some degree of progressivity) and the introduction of a new income tax scale—which, however, has minimal budget effects. More important are the measures on the expenditure side, which included initially a 7% reduction in the budget of the public investment programme (and further reductions in an ad hoc fashion more recently, reaching a drop of over 40% in 2011 compared to the 2009 figure), various cuts in social transfers and benefits, to a value of well-above 5% and more significant cuts in pensions and the public sector. The latter include: a nominal freeze in pensions and public sector wages until 2012; abolishing across the public sector the so-called 13th (and 14th) salary and replacing it with two flat payments of €500 (€400 for pensions); a variable reduction in benefits in the so-called ‘narrow public sector’ (mainly civil servants), ranging from 8% for earnings below €14,000 pa to 13% for earnings over €27,000 pa, representing on average a 10% reduction in nominal pay; a horizontal 10% salary cut for employees in the so-called ‘wider public sector’ (utilities and other state owned enterprises and public bodies) which was later extended and made more progressive; a ‘five out–one in’ rule for hiring in the public sector and abolition of fixed-term contracts; and, prospectively, compulsory dismissals in parts of the ‘wider public sector’ and in local government. Savings from the rationalization of expenditures are also envisaged (by improving public management, rationalizing health expenditure, consolidation of local authority budgets and reduced military procurements) as are increased revenues from tackling tax evasion (although the latter was removed from the Medium-Term Fiscal Strategy Framework (MFSF) as the ineffectiveness of the Greek government in this front made budgeting for savings from tackling tax evasion elusive).

Crucially, with this policy approach, the strategy for stimulating growth as an exit-route out of the crisis has been left to reforms aiming at market liberalization and to wage-depression aiming at restoring international competitiveness—while public consumption and investment are being significantly retrenched. This needed not be the only policy option—but, at least until very recently, it very much appeared to be so given two very real constraints. On the one hand, the evident inability of the Greek government to mobilize resources either internally or from abroad.4 On the other, the unwillingness of the EU to address the crisis in a more holistic and systemic way. Moreover, as the possibility of an EU-induced fiscal stimulus is effectively ruled out, some less drastic policy instruments that could help with stimulating internal demand were also discarded, at least until the end of June 2011. Firstly, a front-loaded absorption of Cohesion Funds with a temporary waiver on the requirement for national co-financing. Such an option was considered by the European Commission in the early stages of the crisis but, largely due to fears of compromising the credibility of Greece and of EMU towards the markets, has been subsequently abandoned (Brunsden, 2009). It was only on 23 June 2011 that the European Commission brought this issue back on the table—and it now seems that a partial release of such funds with reduced Greek co-financing will take place in the second half of 2011.5 Secondly, the provision by the European Investment Bank of specially designated loans to ‘pre-finance’ the national contribution to EU funds—a vehicle which maintains performance incentives while removing the acute budgeting constraints. Greece was offered a €2bn loan under this scheme in 2010, but owing to poor absorptive and administrative capacities has so far opted against this route to dealing with the crisis.6 It remains to be seen how and by how much this will be rectified in the future.

The geographical dimension of the austerity measures

To examine the impact that the measures will have on the Greek regions, we focus on three broad categories: changes in public expenditures (income transfers and public investment), changes in public sector employment and pay and changes in direct and indirect taxation (including measures aiming at tackling tax evasion).

Public expenditures

As mentioned already, in 2010–2011 public investments have been reduced well in excess of the headline 7%.7 Moreover, as the ‘troika’ pressures Greece to accelerate its public investment programme, funds are shifted towards ‘soft’ interventions (for example, on entrepreneurship than on infrastructure), which, however, tend to follow more the distribution of population (than the inverse of regional incomes) thus favouring the main urban areas. Concentration in favour of the same areas also seems to increase for infrastructure investments as some projects in the periphery are being sacrificed.8 As a result, peripheral areas that have traditionally relied more on ‘hard’ public investments may lose out dearly. On the basis of past allocations (for the period 2005–2008—see Figure 1), the worst affected regions seem to be those of Western Macedonia (which, with public investment representing 5% of local GDP, may lose more than 0.5% of its GDP in foregone public investment), Ipeiros and the North Aegean. In contrast, in regions such as Attica, Athens, Crete and Thessaloniki, the impact will be minimal (closer to 0.05% of local GDP).

Figure 1.

Public investment and income transfers by region.

Note:Regions are grouped along four quartiles (darker shades represent higher values). Public investment data (share of regional GDP) refer to average 2005–2008 values and are derived from Monastiriotis and Psycharis (2011). Income transfers (state benefits as a share of average household incomes) are derived from the 2004–2005 Greek Household Budget Survey (ELSTAT).

Figure 1.

Public investment and income transfers by region.

Note:Regions are grouped along four quartiles (darker shades represent higher values). Public investment data (share of regional GDP) refer to average 2005–2008 values and are derived from Monastiriotis and Psycharis (2011). Income transfers (state benefits as a share of average household incomes) are derived from the 2004–2005 Greek Household Budget Survey (ELSTAT).

Similarly, on the basis of the most recently available household income data (right panel of Figure 1), the effects of the cuts in benefits and other income transfers to households, which are in the area upwards of 5%, seem also to affect more strongly regions in the northern and western periphery (as well as, in this case, Crete and Thessaly). Attica, Athens and the South Aegean remain the least affected regions. Again, the effect is projected to be a disproportionate decline in incomes in the periphery, with East Macedonia and Thrace being by far the most affected region (experiencing a projected decline by over 0.5% of household incomes).

Public sector cuts

Despite their notable spatial variation, however, these effects are not particularly sizeable, relatively speaking. Indeed, the main effects on private consumption and household incomes are anticipated to come from the significant reductions in pensions and public sector pay. Using salaried income data from the Greek Labour Force Survey and data on salaried and total household income from the Greek Household Budget Survey (HBS), we calculate that before the crisis, the public sector accounted for close to 20% of total disposable household incomes in the country, while another 20% was accounted for by pensions. On the basis of this, the cuts already implemented and the additional prospective cuts for high earners in the public sector amount to an income reduction of over 4% nationally. Adding to this, the impact of the public sector employment cuts (almost universal abolition of fixed-term contracts, ‘5 out–1 in’ rule, and further downsizing of employment in public utilities) suggests a much more significant effect than that of the cuts in pubic expenditures (in static terms—not accounting for possible multiplier effects). As is shown, however, in Table 1, this effect can be particularly uneven across space. Combined public sector pay and pensions constitute close to or over 50% of household incomes in the north and north-west of the country (Ipeiros, Western Macedonia and North Aegean), while they are less than 35% in South Aegean and Crete (close to 40% in Athens and Central Greece). Assuming a similar geographical allocation of cuts, this implies a reduction in household incomes by some 40% more in the northern periphery than in the south.

Table 1.

Selected components of household income by region

Region All public sector (%) Pensioners (%) All affected incomes %) Public sector temps (%) Central government high wages (%) Public utilities (%) Projection of total effect (%) 
North-west 
    West Macedonia 28.40 20.30 48.70 2.70 1.90 9.20 7.97 
    Ipeiros 23.80 29.70 53.50 1.60 2.20 1.00 6.90 
North and north-east 
    East Macedonia and Thrace 15.10 20.20 35.30 0.70 1.30 0.30 4.25 
    Central Macedonia 16.30 22.70 39.00 1.10 1.20 1.10 4.96 
    North Aegean 22.80 27.10 49.90 1.50 2.40 2.10 6.61 
Western 
    Ionian Islands 16.90 27.00 43.90 1.00 0.30 0.80 5.22 
    Western Greece 16.90 27.10 44.00 0.90 1.90 0.80 5.40 
Central 
    Thessaly 19.50 22.10 41.60 1.20 1.10 0.90 5.26 
    Central Greece 17.60 22.40 40.00 1.10 1.10 1.70 5.11 
    Attiki 18.80 20.50 39.30 1.00 1.30 1.60 4.99 
South 
    Peloponnese 16.00 24.40 40.40 0.80 1.30 1.90 4.99 
    South Aegean 16.30 18.60 34.90 1.40 0.00 0.80 4.55 
    Crete 12.80 21.90 34.70 1.00 1.40 0.50 4.43 
Metropolitan 
    Athens 18.80 20.50 39.30 0.70 1.80 1.70 4.86 
    Thessaloniki 16.30 22.70 39.00 0.60 2.50 0.70 4.77 
Region All public sector (%) Pensioners (%) All affected incomes %) Public sector temps (%) Central government high wages (%) Public utilities (%) Projection of total effect (%) 
North-west 
    West Macedonia 28.40 20.30 48.70 2.70 1.90 9.20 7.97 
    Ipeiros 23.80 29.70 53.50 1.60 2.20 1.00 6.90 
North and north-east 
    East Macedonia and Thrace 15.10 20.20 35.30 0.70 1.30 0.30 4.25 
    Central Macedonia 16.30 22.70 39.00 1.10 1.20 1.10 4.96 
    North Aegean 22.80 27.10 49.90 1.50 2.40 2.10 6.61 
Western 
    Ionian Islands 16.90 27.00 43.90 1.00 0.30 0.80 5.22 
    Western Greece 16.90 27.10 44.00 0.90 1.90 0.80 5.40 
Central 
    Thessaly 19.50 22.10 41.60 1.20 1.10 0.90 5.26 
    Central Greece 17.60 22.40 40.00 1.10 1.10 1.70 5.11 
    Attiki 18.80 20.50 39.30 1.00 1.30 1.60 4.99 
South 
    Peloponnese 16.00 24.40 40.40 0.80 1.30 1.90 4.99 
    South Aegean 16.30 18.60 34.90 1.40 0.00 0.80 4.55 
    Crete 12.80 21.90 34.70 1.00 1.40 0.50 4.43 
Metropolitan 
    Athens 18.80 20.50 39.30 0.70 1.80 1.70 4.86 
    Thessaloniki 16.30 22.70 39.00 0.60 2.50 0.70 4.77 

Note: Shares show income generated by the specific category as a proportion of total disposable household income. Data are derived from the 2004 and 2005 waves of the Greek Quarterly Labour Force Survey and the 2004/2005 Household Budget Survey (ELSTAT), based on author’s calculations. The projection of the total effect (last column) is based on the following calculation: 20% cut in public utilities pay plus 80% cut in fixed-contract incomes plus 25% cut in high-wage incomes in central government plus 10% cut in pensions and in the remaining public sector.

The effects, however, can be even more pronounced owing to the composition of public sector employment in each region. The three most affected regions in the north and west of the country also possess by far the highest shares of incomes generated by fixed-term contracts in the public sector—where the cuts are even more drastic as such jobs are simply being lost. Together with the regions of Thessaloniki and Western Greece (again in the north and west of the country), these three regions also posses the highest shares of incomes accounted for by high-salary earners in the public sector (for example, 2.2% in Ipeiros versus 1.3% in the Peloponnese), who also experience more severe cuts.9 Of them, the North Aegean and especially Western Macedonia also have unusually high shares of incomes generated in public utilities, which are also disproportionately affected. All in all, the three regions that stand out to be more severely affected by the public sector cuts are those of Western Macedonia, the North Aegean and Ipeiros—which also have the lowest shares of private sector employment and the weakest industrial bases. According to the figures presented in Table 1, the measures taken in the public sector can induce a negative income effect of between 6.5 and 8.0% in these regions, which contrasts vividly with the estimated effect in the southern and metropolitan parts of the country, which is projected at closer to 4.5%.

Taxation

We start here with the changes in indirect taxes, relying on national level estimates of the effects of these changes by income decile (from Matsaganis and Leventi, 2011) combined with geographical information on the distribution of household incomes from the HBS. Unsurprisingly, the incidence of low incomes, for which rises in indirect taxation constitute a greater erosion of disposable incomes, is highest in the same regions previously projected to suffer most from the public sector cuts (see left panel of Figure 2). Nationally, without taking into account the additional measures included in the MFSF in June 2011, Matsaganis and Leventi (2011) predict a drop of purchasing power for the median household of about 4.5% (10% for incomes at the bottom decile and less than 3% for incomes at the top decile)—or about 3.8% of average household incomes. Simply projecting these estimates to the regional shares of incomes falling inside each decile of the national distribution of household incomes produces projections for the drop in purchasing power which range from 3.5% in Attica to 4.3% in Ipeiros and above 4% in North Aegean, Western Macedonia, East Macedonia and Thrace, Western Greece and the Peloponnese. The effect is again smallest in the metropolitan, central and southern regions.

Figure 2.

Income shares to total regional household income.

Note: Regions grouped along four quartiles (darker shades represent higher values). All data are from the 2004/2005 Greek Household Budget Survey (ELSTAT).

Figure 2.

Income shares to total regional household income.

Note: Regions grouped along four quartiles (darker shades represent higher values). All data are from the 2004/2005 Greek Household Budget Survey (ELSTAT).

Given the geographical distribution of household incomes, a progressive income tax, in contrast to the effects of indirect taxation, would seem able to counterbalance some of the disproportionality of the effects observed previously. In their analysis of the redistributive effects of the new income tax scales, however, Matsaganis and Leventi (2011) find overall very minor effects, partly owing to the overall reduction in taxable incomes nationally. Despite some more progressive measures in income taxation in 2011,10 the recent reduction in the non-taxable income threshold will make income taxation even less progressive and hurt especially the low-income households. As such households are disproportionately located outside the metropolitan and southern regions, the effect of the new income tax appears to lack also ‘spatially’ progressivity.

Measuring the geographical impact of the attempts to tackle tax evasion is much more challenging. The only available estimates at the sub-national level in Greece come from another recent micro-simulation study (Matsaganis and Flevotomou, 2010), which found tax evasion to be highest in southern mainland Greece (at 16%) and lowest in the large metropolitan area of Athens (at 5.6%)—with northern mainland Greece and the island regions all ranging between 12 and 14%.11 On the face of this, assuming for illustration purposes a 50% success rate in taxing undeclared incomes at an effective tax rate of 35%, the government efforts will reduce disposable household incomes by between 1% (in Athens) and 2.5% (in the rest of the country, including the north-western periphery). If, however, as the government emphasizes, efforts to curb tax evasion concentrate on particular occupational categories (private doctors, lawyers and other professionals) and especially on headline cases in the Greek capital, the results may be drastically different. Figure 2 (right panel) depicts the spatial distribution of the regional income shares generated by self-reported high-income own-account workers. As can be seen, such incomes are largely concentrated in central and southern Greece, including Athens. Was the government to concentrate its efforts to undeclared incomes from this group, the spatial effects of this would run in the opposite direction of most of the spatial effects examined above. As it turns out, however, the government’s efforts on this front seem to be subsiding, not intensifying, after the developments of June 2011.

Does it matter? Exploring longer run threats

The finding that the austerity measures may have spatially heterogeneous income effects, affecting the least developed regions most, does not necessarily imply that policy measures should be taken to correct for this. If cross-regional equilibration mechanisms operate efficiently, the effects of these measures, no matter how asymmetric, will eventually diffuse more or less evenly across space. Locally, reduced public investment, lower disposable incomes and declining (public sector) employment constitute a negative demand shock that will raise unemployment. High unemployment can lead to outmigration (towards less-affected regions—assuming vibrant labour mobility) and a reduction in wages and prices (assuming no price rigidities). The latter will attract investments from other regions where prices and wages remain relatively high thus stimulating demand and creating new jobs. Inversely, outmigration will increase labour supply in the wealthier regions, while capital outflows may reduce labour demand there, putting upward pressures on unemployment (and downward on wages). Assuming, for the sake of the argument, no differences in amenities, preferences, etc, this would result in equalization of wages and unemployment rates across space.12

It is, however, difficult to argue that such equilibration mechanisms operate sufficiently in Greece. Regional working-age migration in the country is extremely low, estimated at 0.5% in 2007 (Monastiriotis and Pissarides, 2011). As already mentioned, industrial production is very concentrated and the regional economies show little diversity in their economic bases. Self-employment (mostly in family businesses and geographically-bound closed professions) represents almost 40% of total employment, while the share of dependent salaried employment outside the public sector is a dismal 30%. Moreover, with manufacturing and energy representing jointly a mere 12.5% of gross value added nationally and only two regions having shares above 20% (Western Macedonia, due to the publicly owned energy plants located there, and Central Greece), the majority of goods and services consumed in Greece are either imported (goods) or locally produced (services).13 Under these conditions, it seems unsurprising that regional unemployment differentials in the country are high and highly persistent (Monastiriotis, 2008)14—and rather natural to expect that the effects of region-specific shocks remain largely localized. With regard to the Greek austerity measures, this implies that these may produce important equilibrium changes in economic activity across the Greek regions, besides their compositional effects examined above. The remainder of this section traces this possibility, not as a means to predict future economic developments in the country but rather as a theoretical exercise aiming at identifying possible cumulative effects that may be triggered under the severity (section 2) and apparent asymmetry (section 3) of the negative demand shock induced by these measures.

Our starting premise is that in the context of acute resource/budget constrains nationally, the negative demand shock, which adds to an existing recession, creates conditions of demand deficiency. On the one hand, rising unemployment in the most heavily affected regions cannot be addressed through new indigenous investment and job creation (even with unemployment raising technically the marginal product of labour) as lower incomes weaken disproportionately the local consumption base. On the other hand, as capital becomes scarcer nationally (due to the substantial rise in borrowing costs), it is improbable that sufficient amounts of capital will flow into these regions from the more developed ones, to take advantage of the reduced costs and rising unemployment there.15 Weakened demand and rising unemployment, especially in the relatively high-skill public sector, may instead create tendencies for outmigration of a brain-drain type, thus lowering productivity in these regions and evaporating any investment incentives accruing from unemployment.16 A Myrdalian-type circular causation effect may well kick-in, at least in the most heavily affected regions, where internal demand recedes the most.

Rational economic behaviour under heightened risk may add to this circular effect. It is well-established in the finance literature (see Eeckhoudt and Schlesinger, 1994; Kimball, 1990) that increased income risk raises financial prudence, leading to a ‘disproportionate’ decline in risk-taking (nationally).17 Recently, Broll et al. (2010) have followed the spatial implications of this, showing that rising financial prudence leads to greater concentration of private capital so that investments are redirected to areas of high agglomeration, even if risks rise ‘evenly’ across space.18 For Greece, this suggests that as the austerity measures intensify economic contraction and uncertainty, ‘prudent’ businesses will cut down on their investments in areas of low demand, weak physical connectivity, thin labour markets and poor infrastructure—even irrespective of the actual size of the negative demand shock experienced in each region—and in contrast concentrate their investments in the main metropolitan areas and especially in the Capital, where large segments of the population, as well as of political power, reside. Thus, investment in peripheral areas will decline further.

Similarly, with rising unemployment nationally, mobile (skilled) workers will have an incentive to concentrate in the big urban agglomerations to benefit from the larger pool of jobs available there. In fact, migration towards the main urban centres may increase even if unemployment rises faster there than in the periphery as long as disinvestment and subsiding demand reduces productivity faster in peripheral areas.19 As a result, economic activity will become increasingly concentrated in better-off high-agglomeration areas. To the extent that this does not happen, for example, due to housing market or other rigidities limiting the geographical mobility of labour, the result will not be cross-regional equilibration but rather higher, and persistent, unemployment in the peripheral regions.

If such a circular causation mechanism is put in place, then a further weakening of the economic potential of the less prosperous regions can follow in a relatively short period of time—in a way that, even if demand differences are restored, the cumulative process of regional differentiation may remain. Consistent with the Kaldorian view of cumulative causation (but also with the endogenous growth literature for knowledge rather than demand-generated spillovers), a drop in the mass (Angeriz et al., 2008) or density (Ciccone and Hall, 1996) of economic activity in these regions will lead to an increasingly slower rate of productivity growth. Slower productivity growth will lead to a relative reduction in economic efficiency and in private returns (wages and profits), thus reinforcing the tendency for out-migration (brain-drain) and disinvestment. As a result, growth differentials between the better-off and the less well-off regions will tend to become permanent, even if the initial conditions that generated them (that is, the austerity measures) disappear.

Of course, the extent to which a negative demand shock in any given region translates into a more structural demand deficiency, with cumulative consequences, depends at least partly on the economic resilience of this region (Pike et al., 2010). More diversified regions and those specializing in products of national or international comparative advantage will be better positioned to overcome the negative effects of the national austerity measures. In the context of the Greek economy, this adds another reason to believe that the impact of the austerity measures may have a cumulative effect of the type discussed above. With depressed demand keeping wage and price inflation nationally at low levels (barring the temporary effects of tax rises on inflation), thus producing effectively an internal devaluation, the regions specializing in tourism and manufacturing exports (mainly Athens, Thessaloniki, the South Aegean and Crete, that is, those least affected by the composition effect) may benefit from an external stimulus (increased exports and international tourist arrivals).20 In contrast, in regions with weak export bases, an internal devaluation will be felt more as an increase in relative import prices—thus further strengthening the circular mechanism discussed above.21

Overall, in the scenario presented above, a clear disparity can emerge between the higher income and less affected economies of the south and the less dynamic and more heavily affected economies of the north. Of course, if such a scenario materializes, it will not be solely the outcome of the geographical asymmetry of the austerity measures but rather of the interaction between this asymmetry and the regional imbalances and weak equilibration mechanisms that characterize Greece. The point made with this scenario is that, with collapsing demand nationally, an additional negative shock directed to the least competitive and perhaps least resilient regions may trigger a cumulative process of divergence, with investment and external demand concentrating in the more extrovert and dynamic regions of the south, especially in and around Athens, at the expense, perhaps even in absolute terms, of the more heavily affected periphery. In such a case, Greece’s regional imbalances will become even more acute, putting serious questions on the country’s ability to achieve balanced and sustainable growth even after it recovers form its current crisis.

Conclusions

An embedded north-south asymmetry in the Eurozone, together with the chronic misreporting of, and lack of prudence in, public finances in Greece, have led to an unprecedented fiscal crisis in the country as the global financial crisis unfolded. Threatened by a seemingly inevitable default, the country is obliged, counter-intuitively, to implement a series of austerity measures that come to add dearly to the recession already experienced by the economy. The situation does not afford Greece the luxury, or time, to devise measures that will address issues of regional imbalance and spatial fairness. Despite that in this paper, we have argued that the spatial implications of the austerity measures may be too big and may have too structural a character to be ignored—even at the current conjunction.

The revenue-generating measures introduced in 2010, as well as the recently announced reduction of the taxable income threshold, appear to have strong spatially regressive effects—besides, or exactly because of, their regressively redistributive character. The new income tax scales appear to do very little to correct for this—at least in the context of declining incomes across the entire distribution—while tackling tax evasion which, if appropriately targeted, could also help in this direction has not progressed nearly as deeply as would be necessary. The expenditure cuts, which to some extent come to substitute for the weaknesses in tax collection, end up also affecting disproportionately the most vulnerable regions and income groups. On the basis of the pre-crisis data used here, three groups of regions can be identified in relation to the direct compositional effects of the austerity measures. Some northern and north-western regions are out to lose the most, with a projected reduction in real disposable incomes (accounting also for the impact of indirect taxation) of well above 10%. Other peripheral and less developed regions of the country (including the non-metropolitan parts of Central Greece and Central Macedonia) will probably experience a negative shock closer to 8–9% of disposable incomes. Finally, the more central and high-income regions of Attica, Thessaloniki, Crete and the south Aegean will experience a significantly smaller shock, perhaps in the area of 6–7%.

The overall effect of this differentiation will be an amplification of existing inequalities. Owing to weak cross-regional adjustment mechanisms and pre-existing imbalances, it is possible that under specific conditions the asymmetry of these effects will trigger a cumulative process of divergence and further regional differentiation that may be hard to rectify also in the future. Of course, whether such a process materializes depends on a variety of other factors, including developments in other policy fronts and in the international economy. The analysis in this paper is not suited to provide conclusive evidence that such a process will indeed be activated. Moreover, the present analysis does not consider important secondary effects (for example, through regional income multipliers or other adjustment mechanisms) that may indeed restrain the cumulative causation forces and instead push towards equilibrium, at least in the longer run. This is certainly a caveat of the present analysis. But although we have no way of testing the empirical validity of the predictions emanating from our analysis, we contend that these do not appear to be particularly implausible. In this sense, devising regionally sensitive austerity policies, even if this appears a tough call for policy in the current climate, may be essential for avoiding the introduction of new distortions to the spatial distribution of economic activity in the country.22

Policy measures that can prevent such potential distortions, by reducing the spatial heterogeneity of the austerity-induced demand shock while remaining consistent with the fiscal consolidation programme, can in fact be found. Domestically, this would require a shift of the fiscal consolidation efforts towards raising revenues from income taxation (of a much more progressive character), which can be fairer not only in a geographical but also in a societal sense. Given the scale of tax evasion in the country, however, the government may find it difficult to raise revenues from this source even with a more progressive income tax. Raising tax rates for high-income households may simply incentivize tax evasion—while pay-cuts and downsizing in the public sector seem already to have contributed to a relaxation of on-the-ground efforts to restore tax revenues and tackle tax evasion (Alpha Bank, 2011). In this context, an externally supported fiscal stimulus, that will help ease the recessionary consequences of the austerity measures and be directed to regions and sectors that are most heavily affected by their horizontal implementation, may in fact be a rather perceptive policy proposal. Interestingly, recent political developments seem to make this policy option also less unlikely. Proposals for the easing of the flow of Cohesion Funds (by relaxing the conditions for co-financing or through direct pre-financing of the national contribution by the European Investment Bank) and for a mechanism for national debt restructuring inside the Eurozone, which were pushed aside in the early phases of this crisis (Brunsden, 2009), seem to be gaining more currency in the current policy agenda (Strupczewski, 2011).

Still, attention to the spatial dimension is largely missing. The Greek government and the EU-IMF-ECB troika continue to show little, if any, attention to the geography of the austerity measures and emphasis remains centred on inducing structural reforms not stimulating growth through public investment. Adherence to the painful and horizontal austerity measures that Greece has committed to is still seen as a major condition for restoring Greece’s (and EMU’s) credibility against the markets.23 It remains to be seen whether policymaking, both in Greece and in Europe, will manage to find a balance between the two objectives (credibility – recovery) in a way that does not compromise the economic prospects and socio-spatial cohesion of the country.

I am grateful to Panos Tsakloglou, Manos Matsaganis, George Petrakos, Christos Koutsambelas and Maria Tsiapa for their help with data collection and to Alejandra Castrodad-Rodriguez, Eleni Xiarchogiannopoulou, John Parr, Theodore Panagiotides and participants at the 2011 Meeting of the Urban and Regional Economics Study Group, the 2011 Conference of the Greek Regional Science Association and to the Yale University Hellenic Studies Programme seminar series for useful discussions and feedback. The useful comments of three anonymous referees are also acknowledged. Full responsibility for opinions expressed and any errors of interpretation remain with the author.

References

Alpha Bank
Weekly Bulletin of Economic Developments
 , 
2011
 
(April Issue), Alpha Bank. Available online at http://www.alpha.gr/files/infoanalyses/weekly20110428.pdf. [Accessed 5 September 2011]
Angeriz
A
McCombie
J
Roberts
M
Estimates of returns to scale and spatial spillovers for EU regional manufacturing
International Regional Science Review
 , 
2008
, vol. 
31
 (pg. 
62
-
87
)
Antoniades
A
Monastiriotis
V
Kalyvas
S
Pagoulatos
G
Tsoukas
H
Reform that! Greece’s Failing Reform Technology: beyond vested interests and political exchange
From Stagnation to Forced Adjustment: Reforms in Greece, 1974–2010
 , 
2011
London
C. Hurst Publishers
Arghyrou
M
Kontonikas
A
The EMU sovereign-debt crisis: Fundamentals, expectations and contagion, European Economy Economic Papers No436
 , 
2011
Brussels
European Commission
Broll
U
Rolda-Ponce
A
Wahl
J
Spatial allocation of capital: the role of risk preferences
Spatial Economic Analysis
 , 
2010
, vol. 
5
 (pg. 
389
-
398
)
Brunsden
J
Commission in Bid to Ease Regional Aid Rules
 , 
2009
 
Article in European Voice (28 May). Available online at: http://www.europeanvoice.com/article/imported/commission-in-bid-to-ease-regional-aid-rules/64968.aspx [Accessed 5 September 2011]
Christopoulos
D
Tsionas
E
Convergence and regional productivity differences: evidence from Greek prefectures
Annals of Regional Science
 , 
2004
, vol. 
38
 (pg. 
387
-
396
)
Ciccone
A
Hall
R
Productivity and the density of economic activity
American Economic Review
 , 
1996
, vol. 
86
 (pg. 
54
-
70
)
Eeckhoudt
L
Schlesinger
H
Increases in prudence and increases in risk aversion
Economics Letters
 , 
1994
, vol. 
45
 (pg. 
51
-
53
)
Fotopoulos
G
Kallioras
D
Petrakos
G
Spatial variations of Greek manufacturing employment growth
Papers in Regional Science
 , 
2010
, vol. 
89
 (pg. 
109
-
133
)
Kimball
M
Precautionary saving in the small and in the large
Econometrica
 , 
1990
, vol. 
58
 (pg. 
53
-
73
)
Martin
R
The local geographies of the financial crisis: from the housing bubble to economic recession and beyond
Journal of Economic Geography
 , 
2010
, vol. 
11
 (pg. 
587
-
618
)
Matsaganis
M
Flevotomou
M
Distributional Implications of Tax Evasion in Greece
 , 
2010
London
Hellenic Observatory
 
GreeSE Paper No31
Matsaganis
M
Leventi
C
Monastiriotis
V
Modelling the distributional implications of the crisis in Greece
Austerity, Recession and Paths to Recovery
 , 
2011
London
Hellenic Observatory
 
GreeSE Papers Special Issue, 5–43
Mitsopoulos
M
Pelagidis
T
Understanding the Crisis in Greece: from boom to bust
 , 
2011
Basingstoke
Palgrave Macmillan
Monastiriotis
V
Hardy
S
Bibby-Larsen
L
Unemployment persistence and economic duality in Greece
Working Regions
 , 
2008
 
Conference Proceedings, Regional Studies Association, Seaford
Monastiriotis
V
Examining the consistency of spatial association patterns across socio-economic indicators: an application to the Greek regions
Empirical Economics
 , 
2009
, vol. 
37
 (pg. 
25
-
49
)
Monastiriotis
V
Internal Haircut or a Driver for Growth?
 , 
2011
 
newspaper Kathimerini (15 April)
Monastiriotis
V
Jordaan
J
Does FDI promote regional development? Evidence from local and regional productivity spillovers in Greece
Eastern Journal of European Studies
 , 
2010
, vol. 
1
 (pg. 
174
-
192
)
Monastiriotis
V
Pissarides
C
Institutional and Non-Institutional Measures of Labour Market Flexibility in Greece, Final Project Report to the Greek Ministry of Finance
 , 
2011
London
Hellenic Observatory
Monastiriotis
V
Psycharis
Y
Spatial and Functional Analysis of the Regional Allocation of Public Investment in Greece
 , 
2011
London
Hellenic Observatory
 
GreeSE Paper No49
Monokroussos
P
Medium-Term Fiscal Strategy (2012–2015), Greece Macro-Monitor (18 April)
 , 
2011
Athens
Eurobank EFG
Petrakos
G
Psycharis
Y
Regional Development in Greece
 , 
2004
Athens
Kritiki Publishers
 
[in Greek]
Petrakos
G
Saratsis
Y
Regional inequalities in Greece
Papers in Regional Science
 , 
2000
, vol. 
79
 (pg. 
57
-
74
)
Pike
A
Dawley
S
Tomaney
J
Resilience, adaptation and adaptability
Cambridge Journal of Regions, Economy and Society
 , 
2010
, vol. 
3
 (pg. 
59
-
70
)
Rowthorn
R
Combined and uneven development: reflections on the north-south divide
Spatial Economic Analysis
 , 
2010
, vol. 
5
 (pg. 
363
-
338
)
Servera
M
Moschovis
G
Tax Shortfalls in Greece, ECFIN Country Focus. vol. 5, No5
 , 
2008
Brussels
European Commission
Skouras
S
Christodoulakis
N
Electoral Misgovernance Cycles
 , 
2011
London
Hellenic Observatory
 
GreeSE Paper No47
Strupczewski
J
EU Executive Wants to Boost Greek Growth with EU Funds
 , 
2011
 
Reuters Online. Available online at: http://www.reuters.com/article/2011/06/21/eu-greece-funds-idUSLDE75K1C420110621 [Accessed 5 September 2011]
1
Consistent with its tradition of ad hoc policy design (Antoniades and Monastiriotis, 2011), Greece has made so far no efforts to undertake an ex-ante impact analysis of its austerity measures (as confirmed in private communications with senior policy officials in the Ministries of Finance and Development).
2
In the first 8 years, since the introduction of the euro Greece is estimated to have experienced, a real currency appreciation of over 20% (Arghyrou and Kontonikas, 2011). To some extent, the EMU design also allowed imperfect monitoring and enforcement of EU rules, which made early action to correct emerging imbalances less likely, thus also contributing to the subsequent crisis.
3
As announced in the Greek MFSF on 15 April 2011. Subsequent revisions in June 2011 brought this figure to above €28bn, as Greece failed to fully meet its 2010 deficit target and GDP continued to decline faster than expected. The revised MFSF introduces additional measures, including a reduction of the taxable income threshold from €12,000–€8000 and the launch of a €50bn privatization programme. The paper does not take into account these additional measures, but focuses on measures introduced in 2010 and until the spring of 2011. It should be noted that the new measures remain spatially horizontal and largely tax-based and in this sense continue to be in the direction of the measures discussed and analysed here.
4
Spreads for Greek bonds remain prohibitive. Government efforts to generate revenues from donations by the Greek Diaspora have proven fruitless. Proceeds from the recently announced €50bn privatization programme are also not directed towards stimulating internal demand as they are already earmarked for buying-back external debt (Monastiriotis, 2011).
6
See http://www.eib.org/about/press/2010/2010-114-eib-bolsters-key-investments-in-greece-with-eur-2-billion-loan.htm?lang=-en [Accessed 5 September 2011]. By June 2011, Greece had advanced only 25% of this loan.
7
According to the Ministry of Finance (Budget Execution Bulletin No6/2011), in the first half of 2011, actual public investments were 42.3% below the budget target.
8
Of the 10 large infrastructure investments (€4.65bn) planned for the next period, 5 concentrate in the metropolitan regions of Athens and Thessaloniki (http://www.enet.gr/?i=news.el.article&id=234965- [Accessed 5 September 2011]).
9
According to current plans, by up to 25% in 2012 (see http://www.enet.gr/?i=news.el.article&id=264850- [Accessed 5 September 2011]). Despite the high shares of high-salary public sector employees observed there, these regions also possess the highest shares of low-paid employment, as discussed later.
10
Specifically, the introduction of a progressive tax-levy in the form of a ‘solidarity contribution’.
11
The differences are mainly compositional, as income under-reporting was found to vary significantly with the type of activity (53% for farmers, 25% for the self-employed and 1% for salaried workers).
12
Diffusion of the local shock can occur also without such factor movements. Unless economic space is deeply fragmented, the local shock in the ‘north’ will also reduce demand for goods and services traded from the ‘south’, thus also reducing incomes and labour demand there.
13
Quite tellingly, the value of imports of goods in 2007 was over double the gross value added of domestically produced goods in agriculture, forestry, manufacturing and energy combined. (Source: ELSTAT, http://www.statistics.gr/portal/page/portal/ESYE/BUCKET/A0702/Other/A0702_SEL30_TS_AN_00_2000_00_2010_01_P_BI_0.xls [Accessed 5 September 2011])
14
In the 10 years before the crisis, the high unemployment regions in the north averaged unemployment rates 3.5 higher than those seen in the south. Although unemployment disparities declined in the early phases of the crisis, more recently unemployment has grown faster in the north (by between 30 and 50% in Western, Central and Eastern Macedonia and Thrace since the third quarter of 2010) and disparities have now returned to their pre-crisis levels. In the spring of 2011, the South Aegean also saw an extraordinary increase in unemployment. (Source:http://www.statistics.gr/portal/page/portal/ESYE/BUCKET/A0101/PressReleases/A0101_SJO02_DT_MM_01_2011_01_F_GR.pdf [Accessed 5 September 2011]).
15
This would require not only perfect functioning markets but also a low degree of financial prudence, as we discuss below.
16
This is especially in view of the fact that public sector employment, which prior to the crisis offered high benefits including ‘a job for life’, constituted a ‘compensating differential’ for the perceivably low amenities offered by non-metropolitan regions in the country.
17
This is the same process that led to the liquidity crisis in 2008, as heightened risk led to a disproportionate reduction in lending.
18
Assume a mobile investor, with an existing investment in a high-agglomeration (low-risk) area, considering a new investment. Risk aversion induces portfolio diversification, pushing the new investment towards a different area, even if this constitutes in absolute terms a higher risk investment. In contrast, financial prudence tends to push investment towards the high-agglomeration area, as diversification of exposure (that is, investing in the second area) entails more risk than a similar investment undertaken in the first area—if not, the investor would not have chosen this area for her first investment. As prudence increases with heightened risks (that is, during crises), the latter effect becomes more dominant and mobile investments tend to move away from low-agglomeration regions.
19
This is consistent both with labour-pooling type Marshallian externalities (where higher job densities imply a higher probability of finding a job for any given level of unemployment) and the Harris–Todaro model of urban migration (where faster productivity growth in urban centres induces migration from the periphery even if the peripheral labour markets clear).
20
It should be acknowledged, however, that the tourism industry is also extremely sensitive to problems of political stability. The recent political upheaval has contributed to a slow-down of tourist arrivals and spending, which is partly reflected in the rising unemployment in the South Aegean.
21
It is unlikely that an internal devaluation will improve demand for the agricultural economies of the north. Agricultural production is highly distorted by subsidies and its structure (small farms, low mechanization) is such that export penetration is constrained more by technology and information problems than by uncompetitive prices.
22
It should be stressed that this is not an argument for regional redistribution per se (for example, through transfers aiming at compensating for long-lasting spatial imbalances) but rather an alert to avoid the amplification of such imbalances by an exogenously induced spatially heterogeneous shock.
23
See the 13 July 2011 news conference by the International Monetary Fund's mission chief for Greece (http://in.reuters.com/article/2011/07/13/greece-imf-idINLDE76C1CU20110713 [Accessed 5 September 2011]).