Deficits, Democracy, and Demographics: Europe’s Three Crises

Abstract

This article argues that the EU and, above all, the eurozone are facing not one crisis – an economic and fiscal one – but three: an economic crisis, a crisis of institutions, and a crisis of demography. These crises are not simultaneous; they are overlapping and self-reinforcing, and there is a high degree of feedback across all three crises. Economically, the euro inflated economic growth and government revenue in the peripheral economies, giving those member states a false sense of their economic prospects. Institutionally, mechanisms were too weak at the EU level to prevent a dangerous escalation of asset (above all house) prices and too fragmented to confront the crisis through an immediate and decisive plan that would provide calm to the markets. Demographically, Europe’s economic and fiscal problems are and will increasingly be exacerbated by the continent’s demographic situation and its projected development, especially in southern Europe.

The post-2007 global economic crisis represents the greatest challenge to the European Union since French President Charles de Gaulle walked out of the European Council in 1965. It does so because it touches on the European Union’s most integrationist project, its ‘big idea’ since the completion of the single market: European monetary integration, including of course the euro.

It comes as little surprise that the bulk of the scholarly and economic-journalistic literature has interpreted the crisis as an essentially economic and, above all, fiscal one. Under this interpretation, the sudden collapse in equity and asset prices had specific causes and effects. It was caused by imbalances within Europe created in large part by northern European countries, above all Germany, that were exporting too much and buying too little. The corollary of this imbalance was the growing current-account deficits of southern European countries, as strong wage and credit growth made their economies increasingly uncompetitive. The effect of the crisis was a sudden plunge in aggregate demand, as consumers tried to deleverage and banks curtailed lending, which led to a rise in unemployment, a fall in tax revenues, and thereby a sharp increase in deficits and debt. Thus understood, the problem suggests a solution: a reflation of the European economy through deficit spending followed by a rebalancing of trade through a targeted stimulation of northern European (and above all German) domestic demand (Blyth 2013b; Hall 2014; Wolf 2013).

This reads like the standard Keynesian prescription for combating a recession, but in this case there is a twist. Advocates of ambitious reflation programmes – both deficit spending and monetary expansion through ‘quantitative easing’ – argue that the 2008 recession was different from all other post-war recessions: the demand shock was so great that it created liquidity-trap conditions in which the zero lower bound on interest rates was reached. In these circumstances, monetary policy has difficulty gaining traction, and fiscal policy becomes even more important (Krugman 2012: 32–5). These unusual circumstances justify, so the argument goes, the acquisition of substantial public debt despite the fact that deficits (in the UK, Spain, Ireland, Greece, and Portugal) and debts (in Italy, Ireland, Greece, Portugal, and even Germany) were (or are) at historic highs. In this interpretation, it was the opposition of certain major players, not least Germany, to major deficits that led to insufficient fiscal reflation and thereby to only tepid economic recovery across Europe (Krugman 2012).

This diagnosis is not necessarily wrong, although there are reasons, outlined below, to treat it with scepticism. Nevertheless, the treatment of the crisis as a chiefly fiscal and economic one obscures the wider context, and that context is both institutional and demographic. In the former, the institutions governing the European monetary union were insufficiently developed to respond to the crisis. Institutional mechanisms were too weak to prevent a dangerous escalation of asset (above all house) prices and too fragmented to confront the crisis through an immediate and decisive plan that would calm the markets (Begg 2009). In the latter, Europe’s economic and fiscal problems are and will increasingly be exacerbated by the continent’s demographic situation and its projected development. For one thing, the costs of unemployment are disproportionately concentrated among the young (in southern Europe), immigrants/ethnic minorities (in northern Europe), and young migrants (across all of Europe). To the extent that this pattern drives youth northwards (and overseas), this will compound the second issue: that ageing populations will mean that some European countries face rising health and pension costs in the context of already very weak fiscal positions and dangerous dependency ratios. In this respect, with the exception of Germany, eastern and southern Europe are set to experience the worst demographic developments. In sum, this means that the southern European countries have some of the most expensive pension systems, the worst demographic trends, and (already) the worst fiscal positions. It is a perfect storm.

Against this background, this article makes two arguments. First, the EU and above all the eurozone face not one but three crises: an economic crisis, as described, but also a crisis of institutions (paradoxically, too weak both in Brussels and in the nation-states) and a crisis of demography. Second, these crises are not only simultaneous; they are overlapping and self-reinforcing. There is a high degree of feedback across all three crises.

Beginning with the first argument, the economic crisis partly occurred because the regulatory edifice surrounding the euro was insufficiently developed. The Stability and Growth Pact was suspended in 2003, and, largely because of German opposition, the Commission had few mechanisms for ensuring fiscal stability and fewer still for encouraging labour market reform. At the member state level, those states with pervasive structural problems – Greece, Portugal, and Italy – lack the mid-level institutions (rules and traditions that govern union–management relations, for example) necessary to build consensus in favour of pension and labour market reform (Gordon 2013; Hall 2014). At the same time, national political institutions are often too weak to push reform through in the absence of (near) consensus (Culpepper 2014). The result is that reforms are only adopted when governments are under great external pressure (whether from the bond markets, the IMF, or Berlin), and, when reform involves sharp reductions in public spending, it disproportionately harms those excluded from southern Europe’s overprotected labour markets: the young, the unskilled, the immigrants (who are generally young and unskilled), and the non-unionised.

At the same time, immigrants and their children become a convenient scapegoat for unemployed and low-income ‘white’ Europeans, a case of outsiders turning on outsiders. Greece’s Golden Dawn is the most extreme example, but far-right parties across Europe (the Front National in France, UKIP in the United Kingdom, the Freedom Party in the Netherlands, the Progress Party in Norway, and People’s Party in Denmark, among others) have used the economic crisis as an opportunity to continue to scapegoat Muslims, immigrants, and asylum seekers (there is naturally overlap among these categories). Europe’s confrontation with large-scale immigration, and above all with Muslim immigration, has undermined support for migration, created deep fears about that migration’s effect on national identity, and re-oriented right-wing politics in Europe towards anti-Islamic sentiment and lobbying (Joppke 2014). It is no surprise that the far-right parties have moved seamlessly from a longstanding demonisation of migrants/Muslims to a demonisation of the very process of European integration. And doing so has produced electoral results: in 2014, the far-right Front National won control of 10 towns in French local elections.

A further and to some degree novel element in recent anti-migrant sentiment and lobbying has been one of the targets: European migrants, particularly those from newly acceded East European member states. Above all in the United Kingdom, but also in Germany and the Netherlands, conservative and far-right politicians have attacked ‘benefit tourism’ (EU citizens allegedly taking advantage of the Union’s free-movement provision to seek higher welfare benefits). As these developments occur in a political context in which member states, notably France, have challenged the Schengen Agreement, it represents a further attack on the legitimacy of the Union.

This article is organised in three parts. It begins by briefly reviewing the economic conditions leading up to the euro crisis. This section critiques both the ‘northern’ argument that southern European/Irish economic problems flowed from excessive public spending and the ‘southern’ argument that an expansion of fiscal policy, stubbornly resisted by northern governments, is the main solution to the crisis. Second, the article explores how institutional weaknesses in the design of European Economic and Monetary Union (EMU) both exacerbated the crisis and prevented Brussels and European member states from combating it. Third, it examines what might be called the political economy of demographics and shows how immigration interacts with economic structures in Europe in a manner that, over the long term, bodes ill for EU member states.

The Euro Crisis: A Brief Summary

Stripped down to its essentials, the euro crisis was a housing crisis that became a banking crisis that became a sovereign debt crisis (Hodson and Quaglia 2009). The collapse in April 2007 of a major US mortgage provider, New Century Financial Corporation, acted as the first trigger and tremor (Hodson and Quaglia 2009: 949), but the causes were deeper (Wolf 2013). Over the previous decade, house prices in the United States and several European countries rose substantially, with particularly high growth in Britain, Spain, Ireland, and France, as well as in the United States itself. House price growth had multiple causes: a cyclical recovery from the sharp falls in the late 1980s and early 1990s, an expanding economy, low interest rates, and more lax credit rules (on, for instance, minimum deposit and maximum income multiples; Carmassi et al. 2009). The last included the now infamous subprime mortgages, which allowed low-income citizens to buy a home worth many multiples of their family income (Begg 2009: 1109). Financial institutions then sold large numbers of these dodgy though attractive loans (as they paid higher interest rates) to other, often European, banks. They accomplished this either by selling the loans directly to other banks or by including them in a package of lower-risk loans that obscured the risks of the bad ones (Authers 2013: 34–5; Blinder 2014: chapter 3). The result was that large numbers of European banks effectively bought into the American mortgage market (Hardie and Howarth 2013). An ever-larger housing bubble resulted, based on the belief that high, even double-digit, percentage increases in house prices could continue indefinitely, or at least that they could not sharply reverse (Blinder 2014: chapter 3).

In 2007, the whole system came close to collapse. House prices in the United States peaked and then started to fall; people began defaulting on their mortgages in increasing numbers, and multiple financial institutions in first the US then Europe were holding huge volumes of worthless debt. Governments began bailing out their banks to prevent a general financial panic. In August 2007, the German government bailed out IKB Deutsche Bank, which was heavily exposed to subprime loans; France’s BNP Paribas suspended three of its investment funds within a few days; and the European Central Bank (ECB) responded to the French move by providing €100 billion in short-term financing to banks. In February 2008, the British government nationalised Northern Rock (whose branches had witnessed 1930s-style queues the previous August) (Hodson and Quaglia 2009: 940–1). The US government let Lehman Brothers go bankrupt on 15 September 2008, but then decided that the risks associated with further bank collapses were too great.

Governments in the United States and the European Union agreed in October of 2008 to provide aid in the form of liquidity support, interbank lending guarantees, and recapitalisation (that is, direct transfers of funds) – all known colloquially as bailouts – to their banks. By the end of 2009, the UK had spent 6.5 per cent of its GDP on bailing out its banks; Belgium had spent 6 per cent; Germany 4.9 per cent; the US 4.9 per cent; the Netherlands 2 per cent; and France and Spain 1.1 per cent (Hardie and Howarth 2013: 12).

These staggering sums were followed by fiscal stimulus programmes designed to flood the economy with money and to save key industries (the automobile industry in the United States and Germany, for instance). In late 2009, the European and American economies appeared to stabilise, but the banking crisis soon became a public debt crisis in southern Europe and Ireland (Krugman 2012: 3). Greece admitted in December 2009 that its total debt was far larger than previously reported – €300 billion, or 113 per cent of its GDP – and a month later it revised its deficit levels up from 3.7 per cent of GDP to 12.7 per cent, a figure still 3 percentage points below the eventual official estimate (IMF 2013: 7; Olivares-Caminal 2011: 3). Whereas deficits were growing across Europe, it was above all in southern Europe and Ireland that they increased dramatically due to a combination of sharp economic contraction, increased welfare spending, stimulus programmes, and the need to bail out domestic banks. In the most striking example, Ireland’s deficit reached 32 per cent of GDP in 2010, mostly because of the bank bailout (Blinder 2014: 411). Bond markets began demanding higher interest rates on southern European and Irish debt. Greece and Ireland teetered on the verge of bankruptcy, meaning they could not afford to cover the debt and interest rate charges required to meet spending commitments. A nightmare scenario seemed possible: Greece would default, and holders of loans in Ireland, Spain, Portugal, and Italy would seek to recover lost capital by calling in the loans, sending all of these countries into a forced default (Authers 2013: 101). If Italy defaulted, France might be next, and the euro would collapse.

The European Union responded in a manner that pleased few. In early 2010, France and Germany announced that euro area member states (EAMS) were prepared to offer €45 billion in bilateral loans to Greece, later raising the figure to €110 million; neither effort calmed the markets (Gocaj and Meunier 2013: 242). Then, in May 2010, the EU launched a major bailout package – €750 billion, made up of €60 billion in EU-backed funds, €440 billion of member state guarantees, and €250 billion of IMF money – and created a new institution, the European Financial Stability Facility (EFSF) (Gocaj and Meunier 2013: 244). Countries that had lost access to capital markets could apply for support; the Commission, with the IMF and the ECB (the ‘Troika’), negotiated the loan, and EU finance ministers had to sign off. Germany was the largest contributor, followed by France and Italy. The mechanism was temporary, and no new loans have been offered under them since 2013. In early 2012, the EU member states agreed to replace them with the European Stability Mechanism (ESM), a permanent mechanism for providing emergency loans (for countries threatened by default and for bank recapitalisation) to member states. Under ESM, lending capacity was raised from €440 billion to €500 billion, interest rates were lowered by 100 basis points, and the ESM enjoys bond buying power in the primary markets (Gocaj and Meunier 2013: 248). In 2012, ESM was given further power to recapitalise banks (Gocaj and Meunier 2013). In the end, German and other euro area member states provided substantial guarantees, but the delay in offering them meant that (i) the EU looked reactive and indecisive, and (ii) the confidence-restoring benefits of the guarantees were more limited than they would have been under an earlier, decisive intervention (Blyth 2013a).

Under the EFSF, large loans were granted to Greece (€246 billion), Ireland (€67 billion), Portugal (€78 billion), and Spain (€41 billion, lower than the €100 billion offered) (Blyth 2013a). In all cases, funds were raised in the bond markets using member state credit ratings as capital, allowing for much lower interest rates than any of these countries could have secured alone. The goal was to reduce these states’ sovereign debt burdens. The European Commission, with strong German steering, insisted on three sets of conditions before it would agree to a transfer.

The first set of conditions involved investor losses on bad Greek loans. After much negotiation and predictable public grandstanding, Greece and the Troika agreed that private investors would be forced to accept a 53 per cent loss on their investments (the ‘haircut’). The second was a rationalisation of the public service and a reduction in public spending in order to bring high public deficits, hovering around 10 per cent of GDP, under control. The public spending cuts demanded were substantial: an 11 per cent reduction in Greece’s deficit within three years; a 9 per cent reduction in Ireland’s within five years; and a 6 per cent reduction in Portugal’s within three years (Hall 2014). Public spending reductions included pension reform, meaning lower benefits and later retirement ages; pensions were often paid at a higher percentage of previous income (or replacement rate) in southern than in northern Europe (to say nothing of North America; OECD undated: 5). The third set of conditions involved the implementation of structural reform, including the easier hiring and firing of workers, more part-time employment, and the liberalisation of competition in certain sectors (see Hemerijck 2012; for the Memorandum for Understanding, see Mitsopoulos and Pelagids 2012: 238–40, 257–9). The Troika and Germany thus attempted to leverage the wealth transfer to press policy reforms that many economists believe are necessary if southern European economies are to achieve non-inflationary and non-debt-driven economic growth (interview with a German Foreign Office official, September 2012).

Outside Germany, and to a degree within it, the chorus of criticism was swift and sharp. The Federal Republic of Germany – as sages from the fields of economics but above all political science and even history informed us – was pursuing a self-defeating strategy (Cohen 2011; Economist 2013; Hutton 2011; Kundnani 2012; Tooze 2012 – the last is a historian). By clamping down on spending, they argued, the result would be less economic growth, lower tax revenues, and higher deficits. If these new deficits led to more austerity, the result would be a vicious circle of low spending, high deficits, and low growth. As Paul Krugman argued in the American context,

when you combine the growing evidence that fiscal austerity is reducing our future [growth] prospects with the very low interest rates on U.S. government debt, it’s hard to avoid a startling conclusion: budget austerity may well be counterproductive even from a purely fiscal point of view, because lower future growth means lower tax receipts. (Krugman 2011; also see Blyth 2013b and, on the US, Summers 2013)

Whatever the precise relationship between spending and growth, in the European context the Keynesian argument suffers from a basic flaw: it suggests that southern European economies can go back to the future. The argument assumes that the relatively high levels of growth experienced from 1999 to 2008 in parts of the European periphery were sustainable. They were not. What parts of southern Europe and Ireland in fact entered was a period of unusually high, credit-driven growth rates. Their governments assumed, as governments tend to do in boom times, permanent spending commitments on the basis of temporary revenues (Mitsopoulos and Pelagidis 2012: 232–3). In Greece and Spain this boost to growth is clearest, but even in countries such as Ireland and Italy where growth decelerated after the euro, growth rates were boosted by unsustainable factors: sharply rising house prices in the former and wage growth in the latter which would later consign Italy to some of the worst post-crisis stagnation (see Figure 1). In effect, the painful consequences of the euro for southern European competitiveness and growth prospects were masked by credit and wage dynamics that would ultimately unravel.

Figure 1 Economic Growth Rates in Select Euro Countries, 1991–2012

Source: OECD. Dates have been chosen to coincide roughly with comparable points in the business cycle (i.e., just before a recession).

A similar pattern emerges in terms of unemployment rates. In Ireland, Italy, Greece, and Spain, unemployment was high – near or above 10 per cent – before those countries adopted the euro (Figure 2; values for Greece not shown due to lack of comparable data). It fell as interest rates converged in the run-up to the euro; continued to fall during the first few years of the euro’s existence; and then rapidly increased from 2007. In Ireland and Spain unemployment levels returned in the 2007–2012 period to roughly 1991–1999 levels; Greece and Portugal in fact overshot. Put simply, if chronically high unemployment characterised Greece, Ireland, Italy, and Spain before the introduction of the euro, then the underlying causes of that unemployment probably had little to do with the euro or even with austerity.

Figure 2 Unemployment Rates in Select Euro Countries, 1991–2012

German growth and unemployment numbers merit comment. Looking only at the euro period, Germany between 2000 and 2006 had an average growth rate of 1.39 per cent, versus 5.89 per cent in Ireland, 4.3 per cent in Greece, and 3.6 per cent in Spain. The reasons were not hard to explain. Germany’s lavish post-1990 public spending on unification predictably resulted in inflation. This was exacerbated by the resulting delay to much-needed structural reforms, and Germany’s quintessentially export-based economy consequently lost much of its international competitiveness (Mayer 2012: 82). By 2000, taxes were higher, labour markets were more regulated, and social security was more expensive in Germany than its competitor countries (Mayer 2012: 83). To make matters worse, Germany introduced the euro at what many regarded as an unjustifiably high exchange rate. Rather ironically, a Germany that is now declared the main beneficiary of the euro was understood from 1999 to 2005 as one of the currency union’s chief losers (Smith 2003). In 2000, Germany suffered a current account deficit – an oft-forgotten fact.

Unable to allow the German mark to drift downwards because of a single currency that was desired more by its partners than by Germany itself, policy-makers responded the only way they could: through reducing labour market regulation and using welfare reform to encourage people back into work. At the same time, German unions and employers, through mechanisms of wage coordination, reduced wage growth through extended work and below-inflation wage increases (Carlin and Soskice 2009: 72).

In a single-currency zone, Germany’s increased competiveness had zero-sum implications for other eurozone members (Hall 2014). As Germany’s competitiveness improved, that of southern Europe worsened. Easy money and a credit-driven boom resulted, predictably, in rising wages and prices. Greece shed 30 per cent of its competiveness from the mid-1990s until 2011, for example, chiefly because of the effect of a loose fiscal policy on wage and price inflation (Alexiou 2011). As Greece, Ireland, Portugal, and Italy voluntarily, indeed enthusiastically, entered the eurozone, they gave up their traditional tool for restoring competiveness: devaluation (Hall 2014). Since a single-currency zone means that some countries will inevitably be less competitive than others, there is a strong case for increased intra-eurozone fiscal transfers, as occurs automatically between US states within federal programmes such as Social Security and Medicare. Yet this is beyond the pale for European citizens, at least at this point.

A comparison of interest rate spreads sheds light on a dynamic similar to that seen with growth and unemployment rates. Figure 3 compares the spreads between select states and Germany between 1994 and 1995, in the aftermath of the last currency crisis (following the near-breakup of the Exchange Rate Mechanism), and in 2010–2012, the current one (following the perhaps less-near breakup of the euro).

Figure 3 Long-term Interest Rates in Select European Countries, 1994–2011

Source: OECD.

We have a similar story: interest rates in the other countries fell relative to Germany’s in the run-up to the euro’s introduction, converged with German levels following the launch of the euro because financial markets assumed that member states would not be allowed to default, and then diverged from German rates during the fiscal crisis, when default(s) seemed a real possibility. The trend began several years before the euro’s introduction – from the mid-1990s – as markets priced in expectations of euro entry. From 1998 to 2006, the GIIPS (Greece, Italy, Ireland, Portugal, and Spain) levels had all but converged with Germany’s. By 2010, they returned to near their pre-euro rate (there is no OECD data on Greece for this period). As Shambaugh (2012: 194) concludes,

a very good predictor of current austerity is the interest rate on long-term debt back in 1993 [the first year of EU-wide data collected by Eurostat]. In other words, if markets were distrustful of a country two decades ago, it is likely that the same country is being forced to undergo austerity today, as the markets seem reluctant once again to lend.

The euro-driven convergence on German bond rates provided weaker peripheral economies with a windfall in the form of low interest rates. To the degree that these countries channelled (or allowed the channelling of) this windfall into private, debt-driven consumption rather than into investments in education and productivity-enhancing infrastructure, into easing the costs of structural reform, or (in Greece and Italy) into reducing high structural debt, they squandered a golden opportunity and left themselves exposed when interest rate spreads increased.

Finally, house prices in the eurozone also followed the same trajectory as interest rates, with admittedly more national variation (Figure 4). Taking the 1997 period as a baseline, Greek, Irish, Italian, French, and Spanish property prices shot up to figures between 1.5 times (Italy) and 2.5 times (Ireland) their late-1990 levels and then fell back to 1.2 (Ireland) to 1.6 (France) times those early levels.

Figure 4 House Prices, 1997–2012 (100 = long-term national price-to-income average)

The picture is thus reasonably clear. In GDP growth, house price growth, employment growth, and interest rate spreads, the euro distorted the European economy. It gave a sudden boost in the form of cheaper credit to southern Europe, the cheaper credit led to a housing and consumption boom, and the housing- and spending-driven growth allowed unemployment levels to fall steadily, in some cases sharply. The euro did not cause the underlying economic problems affecting southern Europe; rather, it allowed the GIIPS to ignore them. Similarly, excessive public spending was not (Greece excepted) the problem in itself; it was rather the impression created by the euro that public spending was not excessive, given existing tax structures and compliance. Scholars have made the first point but overlooked the second. Regression analysis does indeed show that the relationship between the change in public debt from 2001 to 2007 and post-crisis interest rates spread is ‘weak and not statistically significant’ (Shambaugh 2012: 176). This observation, however, overlooks the bigger picture: tax receipts in the European periphery were inflated by the boom. Any permanent spending commitments entered into during that period were made on the basis of temporary revenue. In this sense the euro was, as George Soros (2012) concludes, the bubble. In the case of Italy and Greece, most problematically, new spending commitments were undertaken against the backdrop of already massive public debt – over 100 per cent of GDP.

Institutions

To this point the story has been chiefly an economic one. But the economic crisis has been powerfully shaped by existing institutions, at both the EU and member state levels. And these institutions proved to be woefully inadequate in heading off or in coping with the crisis. Although the housing boom was no secret at the time – commentators in the Financial Times and The Economist frequently warned of the perils of the great upsurge in house prices (Economist 2002, 2003, 2005) – there was little that the EU could do.¹ The Commission had, somewhat unsurprisingly, no authority over minimum down payments for mortgages, maximum amortisation periods, or maximum debt/personal income levels. Moreover, an absence of a fiscal union meant that Brussels could place no formal limits on national public spending (Hall 2014). The German and French governments removed the EU’s formal powers by suspending in 2003 the Commission’s ability to fine countries that exceeded deficit limits specified by the Stability and Growth Pact, upon which Germany had insisted when establishing to the euro.

If the European Union could not prevent the crisis, it was even more ill equipped to limit it once it was in full swing (Begg 2009: 1112; Howarth 2012). Largely because German governments fear the inflationary effect of reflationary measures (on this, see Hall 2014 and Howarth 2012), the ECB lacked the crucial power possessed by other central banks: the capacity to use decisive action, such as bond-buying in the primary markets (prohibited by the Maastricht Treaty) to reassure the markets, often at lesser cost than later moves (Howarth 2012).² It is noteworthy that one of the most influential actions by the ECB in reducing German–GIIPS spreads was rhetorical: a promise by Mario Draghi, head of the ECB, in July 2012 to do ‘whatever it takes’ to save the euro, an intervention that involved no immediate financial outlay. In all other cases, partly because of the euro’s internal decision-making process, and partly because of the need to secure German agreement and German guarantees, the ECB’s actions have been belated and less effective than identical action taken earlier on would have been – particularly action that might have been packaged in resolute-sounding discourse. At the same time, politicians in northern Europe felt unwilling or unable to build a political case in favour of fiscal transfers (with, for instance, the argument that the crisis had lowered interest rates and thus saved German and other northern governments a great deal of money in interest payments). They preferred instead to encourage or tolerate a view held widely among northern publics that southern fiscal profligacy was at fault (Blyth 2013a; Hall 2014). Only the ECB’s capacity to funnel funds to southern Europe through surreptitiously buying GIIPS debt on the secondary bond markets (that is, buying bonds from actors who bought them directly from governments; Mayer 2012) prevented a worse crisis.

The story at the member state level is, to put it mildly, more complicated. When EU member states joined the currency, they gave up the most basic instrument for coping with asset price bubbles: changing interest rates. But they had many others. Member states possessed all the tools that the EU lacked: the capacity to demand higher minimum deposits for mortgages, lower maximum loans, higher capitalisation requirements for the banks, and the prohibition of accounting tricks that allowed banks to overleverage themselves without revealing the matter to their shareholders. Largely because those member states (along with the United Kingdom) were enjoying the prosperity and revenues associated with the boom, neither southern European countries nor Ireland took advantage of these tools. Recognising this fact is important for understanding the crisis, because it highlights the fact that a housing boom was not inevitable given interest rate convergence in the EU (Shambaugh 2012: 173).

National control over setting interest rates and regulating the banking sector do not alone determine a state’s capacity to absorb economic shocks. A further factor is the extent to which intermediate institutions – rules and norms governing the labour market, relations between business, unions, and the state – allowed a government to take the steps necessary to avoid an economic crisis or to cope with it once it occurs (Gordon 2013). In this respect, meso-level institutional arrangements in parts of Europe are highly pro-cyclical: the countries worst affected by the euro crisis – Greece, Portugal, and Italy – are least able to contain economic booms and least able to cope with economic busts (Scharpf 2013).

Welfare and labour market policy are closely related to these problematic institutions (Gordon 2013). Southern European welfare states have three basic characteristics: first, they have expensive pension schemes, as a share of GDP and/or total social spending, privileging the old over the young; second, they benefit insiders to the exclusion of outsiders; and third, they embed welfare arrangements in labour markets that are either inflexible or that impose the cost of flexibility on those on the margins of the market (Featherstone 2013: 735–6; Rhodes 1996; Rueda 2006). Before the post-crisis reforms, Greece had the highest replacement rates for pensions in the OECD, with an average pension level of 93.6 per cent of previous earnings; this level compares with 51.2 per cent in France, 42 per cent in Canada, and 40.05 per cent in Germany (OECD undated: 5). The pension replacement rates for an average worker with a long work history were 73 per cent and 69.3 per cent respectively in Spain and Italy (OECD undated: 5). Portugal was the only country in southern Europe with a relatively modest level of 53.6 per cent (OECD undated: 5).

In terms of employment protection legislation (EPL), Portugal had the strictest set of laws and regulations in Europe, followed by Spain, Greece, and France (OECD 2004: 72). Among southern European countries, only Italy had moderate levels of EPL, clustering with Germany, but this mostly related to its complete liberalisation of temporary contracts; it had little to do with core workers (OECD 2004). All of the countries with traditionally high levels of EPL have high levels of youth unemployment: 53.2 per cent in Spain, 55.3 per cent in Greece, 37.7 per cent in Portugal, 35.3 per cent in Italy, and 23.8 per cent in France (OECD 2013).

Finally, southern European countries suffer from an institutional bias against structural market reform. In several instances, efforts to introduce labour market reforms (easing hiring and firing), product market reform (reducing obstacles to establishing a business), or pension reforms (increasing minimum working ages and reducing replacement rates) crashed against the rocks of opposition of vested interests, particularly the trade union movements (for details, see Carrera et al. 2009; Immergut et al. 2007; Ochel 2008; Rueda 2006).

The result has been a predictable dynamic. When reforms were proposed, insider groups reacted defensively, bringing their members en masse to the street and often blocking reform. Outsider groups which might have benefited from the reform (the young, immigrants, the unemployed) had no strong organised presence, few connections with insiders, and few spokespeople (Featherstone 2013: 740; Gordon forthcoming; Mitsopoulos and Pelagidis 2012: 10). The Greek experience is illustrative. In the late 1990s, Prime Minister Costas Simitis of the Pan Hellenic Socialist Movement (PASOK) attempted to improve four features of the labour market: (i) increasing flexibility in working hours, (ii) restricting overtime in order to share work, (iii) reducing employers’ national insurance contributions for new hires, and (iv) relaxing limits on mass redundancies (Papadimitriou 2005: 388–9). The plan utterly failed in the face of union opposition, and only a bland and ineffective reform measure passed (for details, see Papadimitriou 2005: 389–91). Reforms thus faced a version of James Q. Wilson's (1984) classic ‘concentrated benefits, diffuse costs’ barrier: those who enjoy the concentrated benefits of unfair state provision or support (pensioners, for instance) had a far greater ability and incentive to mobilise than those who paid the diffuse costs (higher unemployment, above all youth and migrant unemployment). Thus, a second effort by Simitis to raise the retirement age and reduce the replacement rate to a still generous 60 per cent led to a general strike and the collapse of the reform as ministers who had indicated initial support reneged on their promises (Featherstone 2013: 744). The retirement age was only raised to 65, and then later to 67, following massive external pressure from Germany and the Troika (Hemerijck 2012: 76–7).

Added to the mix is a temporal issue: the costs of reform, in terms of staunch insider opposition and sometimes in the form of short-term financial costs, are immediate, whereas the gains are medium to long term (Featherstone 2013: 740). Where the capacity of state institutions to determine and implement policy change is low (as in Greece) or where the party system is too weak to build consensus in favour of reform (as in Italy), the chances of reform are highly unfavourable (Culpepper 2014; Featherstone 2013: 738; see also Carrera et al. 2009; Immergut et al. 2007).

The blocking of reform efforts to core institutions forced the pursuit of flexibility towards the non-unionised, immigrant, and illegal labour market, in which the most excluded groups pay the highest price for pursuing competiveness. Thus, the societies with the greatest protections for insiders – Greece, Italy, and Spain – have two features consistent with high insider protections: high youth unemployment and a large, informal (or illegal) economy with relatively low migrant unemployment (on the latter, see Baldwin-Edwards 2013). This is so because many – but by no means all – migrants are incorporated into the informal economy and, given labour market restrictions, the informal economy is the only context in which there can be flexible labour markets. High youth unemployment and a large informal sector thus reflect the same causes: inflexible core labour markets, with a high degree of insider protection, and a weak capacity for reform.

Ireland represents a different type of case from southern Europe and northern Europe. From the late 1980s, in an era of consistently high inflation, the government sought to create a coordinative mechanism, along Austrian/Swedish lines, that was designed to achieve sectoral wage agreements that would secure wage restraint (Boucher and Collins 2003). The strategy worked, and by the 1990s, Ireland had in place the basic conditions for its transformation into the Celtic Tiger: an advantageous corporate tax regime; substantial inflows of efficiently used EU structural funds; a young, educated workforce; and falling unit labour costs from the late 1980s until the late 1990s (OECD 2001: 22).

As the Irish economy grew at a staggering rate – often 10 per cent per year – two things happened. First, Irish house prices ratcheted up at a rate that exceeded even those in the UK, reaching twice their long-term average between 1994 and 2004. Money poured into the market in a speculative frenzy, and both banks and private individuals became massively overleveraged (tradesmen became real estate speculators, purchasing rows of houses on credit). Second, in the face of rapidly increasing private sector wages and wealth, governments bought industrial peace through ‘benchmarking’, which meant pegging public sector wage increases to private sector ones. In 2000, the Irish government established the Public Service Benchmarking Body (PSBB) in the context of labour unrest (Kelly et al. 2008: 3–4). Two years later, it issued a report recommending salary raises for 138 pay grades, ranging from 2 per cent to 27 per cent, with an average of 8.9 per cent (Kelly et al. 2008: 4; Report of the PSBB 2002: chapters 7–12). Eight further above-inflation wage deals followed, and between 2003 and 2006 public sector wages grew by 27 per cent, compared with 17 per cent in the private sector (Kelly et al. 2008: 13). By 2008, Ireland had the highest public sector wage growth in the EU, and the second-highest public sector wage bill as a percentage of general government expenditure (28 per cent) after Portugal (29 per cent) (2004–2008 averages; data from Holm-Hadulla et al. 2010: 9–13). The comparable figures for Germany, Austria, and the Netherlands are 16 per cent, 18.5 per cent, and 19 per cent respectively (Holm-Hadulla et al. 2010: 9–13). It was an example of making permanent, or at least ‘sticky’, spending commitments on the back of temporary revenue. Although Ireland was the only country in the EU that benchmarked, a similar dynamic can be observed in southern Europe: Greece, Spain, Italy, and Portugal have seen the largest increases in public sector wage growth relative to the private sector (Holm-Hadulla et al. 2010: 10). By contrast, such wage growth in Belgium, Germany, France, the Netherlands, Austria, and Finland has been flat (Holm-Hadulla et al. 2010: 10).

The sharp fall in property revenue and the increase in public sector spending would have strained Ireland’s public finances in and of themselves. What ruined them was the decision to provide banking guarantees to banks that, to make matters worse, deliberately underestimated their degree of exposure to bad property debts. The Irish bank bailout is estimated to have cost roughly €70 billion, in an economy with a GDP of €200 billion (Elliot and Treanor 2011). Thus, the bank bailout has single-handedly added some 40 per cent of GDP to its public debt, which went from about 30 per cent of GDP in 2007 to 100 per cent in 2010, or about half of the increase. The other half reflected some combination of the longer-term effects of earlier spending commitments and the sharp drop in tax revenue as unemployment surged.

Demography

The final crisis is demographic, and it has two dimensions: fertility and migration. In the case of fertility, as so often in the European crisis, the cycles appear vicious rather than virtuous. With the exceptions of Germany (which has a northern European economy but southern European fertility rates) and Ireland (which has the only above-replacement birth rate – 2.1 – in Europe), the countries in Europe most affected by the economic crisis are precisely those with the lowest fertility rates. As Figure 5 demonstrates, Greece, Spain, Italy, and Portugal all have below-replacement birth rates, with particularly low rates in the Spanish and Portuguese cases.

Figure 5 Fertility Rates in Select European Countries, 1970–2010

With the exception of Spain (where debt is rising quickly at the time of writing), these countries also have some of the highest sovereign debt in Europe. And they also have, following the south European welfare state model, expensive pension regimes. Under current trends, a simultaneously ageing and shrinking population will lead to a drop in tax revenue as workers retire and to a commensurate increase in public expenditure as they draw pensions and resort increasingly, as the aged do, to public health care. The two trends could set public debt, which is already often at dangerous levels, on a sharp upward trajectory. There is some indication that bond markets’ scepticism about Italy’s creditworthiness is informed by doubts about the country’s ability to cope with demographic change (Blyth 2013b: 70). In northern Europe, Germany has similarly poor demographics, which has raised some doubts about that country’s capacity to continue to act as the core of the European economy (Daley and Kulish 2013). A demographically impoverished Germany, to borrow from Keynes, is a demographically impoverished Europe.³

If the EU is to avoid the worst effects of population ageing and decline, it will need to raise the retirement age still further, increase the female labour participation rate, encourage where it can a higher birth rate (though past efforts suggest this is extremely difficult), and expand immigration. Immigration is not a simple, one-off solution to demographic problems however, as immigrants obviously age like everyone else and their initially higher fertility rates eventually converge with the overall population over the medium term (Kulu 2005). Immigration reforms nonetheless can make population decline more manageable by making it more gradual. A 10 per cent population decline over 50 years, for example, is easier to manage than an identical decline over 10 years. In addition, the more immigrants a country receives, the smaller the decline will be: demographic modelling of the effect of immigration on future population trends shows that (relatively) modest differences in annual intake can have large long-term effects (see the UN Population Division 2000 report). The numbers are in all cases large, though, often much larger than migration-sceptic European publics may be prepared to accept, and under no scenario is the problem of population decline eliminated.

Immigrants, in fiscal terms, are only of use if they work, and large parts of Europe have a particularly poor record of integrating migrants into the labour market (Hansen 2007, 2011). Even before the crisis, unemployment rates among migrants and ethnic minorities (that is, migrants and their descendants) in much of Europe were much higher than they were among the general populations. Figure 6 shows the relative unemployment rate of migrants compared to overall unemployment across much of the EU before and after the fiscal crisis.

Figure 6 Ratio of Foreign-born Unemployment Rate to Native-born Unemployment Rate, 2007 and 2011

Source: OECD. Data is unavailable for Germany in 2007.

As this figure shows, in northern Europe among the foreign-born unemployment is between 2 and 2.5 times the national average. By contrast, in the classic immigration countries of the United States, Australia, and New Zealand and in southern Europe, migrant and ethnic minority unemployment is far lower. The difference between the classic countries of immigration and northern Europe is partly a function of the skills sets of the migrants themselves. Australia and New Zealand (along with Canada) select migrants according to education and skills, and high levels of education correlate with academic success and, therefore, better employment prospects (Bauer and Riphahn 2007; Liebig and Widmaier 2009). By contrast, states in which unemployment among the foreign born is particularly high – Switzerland, Austria, and Belgium – have for decades employed negative selection migration policies: they have heavily selected low-skilled migrants whose jobs have either largely disappeared or who, in the case of many asylum seekers, never entered the legal labour market in the first place.

This is, however, only part of the story: southern Europe also selects negatively, and yet in 2007 its relative unemployment rates for foreign-born residents were below 1.5 and thus notably lower than northern Europe. By 2012, they were still lower but had during the recession increased faster than in northern Europe, with Switzerland being an exception. Both numbers likely reflect a single cause: the greater flexibilisation of migrant labour in southern Europe. In southern Europe, migrant labour is the flexible segment of the labour market, meaning that unemployment falls faster in good times and climbs faster in poor ones (on this, see Triandafyllidou 2008). Migrants were let go for precisely the same reason they were hired: as outsiders, they had little job protection or rights (OECD 2013: 75). They were thus disproportionately employed in the booming construction sector (OECD 2013: 75; on the trade-off between the rights given to migrants and the numbers of them a country and economy can absorb, see Ruhs 2013).

Migrants are Europe’s outsiders. Yet so are its youth. The one group whose employment position has deteriorated even more rapidly than that of migrants in southern Europe is the youth demographic. Figure 7 compares youth unemployment rates before and after the crisis. With the exception of Germany and Austria, youth unemployment increased across Europe, and it skyrocketed in the countries most affected by the crisis. Youth are disproportionately outsiders in Europe: they are concentrated in cyclical sectors and in temporary jobs, and employers face few costs in firing them (OECD 2013: 75). When the young are also immigrants, they face a dual disadvantage: across Europe, employment among foreign-born 15–24-year-olds is much higher than it is among the native born in the same age bracket: 50 per cent higher in Greece and Italy, 100 per cent higher in Germany and the Netherlands (8 per cent versus 17 per cent), and almost 200 per cent higher in Austria (7 per cent versus 21 per cent) (OECD 2013: 75). It is worth noting, once again, that southern Europe produces better migrant outcomes.

Figure 7 Youth Unemployment Rates in Select OECD Countries (per cent), 2007 and 2011

Source: OECD.

In terms of labour market rigidities, European youth – migrant as well as native born – pay a price for the security enjoyed by older workers (OECD undated: 12). And youth unemployment, in turn, further exacerbates the demographic crisis. Between 2008 and 2012, for example, 700,000 people, including both Spaniards and migrants to Spain, emigrated from the country (Fundación alternativas 2013). Since such emigrants are disproportionately young, this migration further exacerbates Spain’s demographic and fiscal challenges by raising the country’s dependency rate and undercutting the revenue-raising capacities of governments. Given the absence of direct fiscal transfers in the eurozone (the institutional dimension), then, this emigration causes problems that would not occur in a federal country such as the US or Germany.

Conclusion

The post-2008 crisis in the eurozone is an economic, institutional, and demographic one, with feedback among all three areas. The euro was introduced in 1999 with the goal of reducing currency swings (particularly against the German mark), expanding trade, and encouraging convergence among European economies. It achieved some of these goals: there has been by definition no currency crisis within the eurozone since 1999, and the euro may have stimulated more trade, though the evidence on this point is inconclusive (Silva and Tenreyro 2010). Rather than encouraging convergence, however, the euro resulted in divergence (Estrada et al. 2013; Hall 2014; Landman 2011): it allowed peripheral economies in Europe to ignore long-term structural problems and to use the sudden boost of lower interest rates and easy credit to underpin consumer-driven booms and longstanding imbalances in tax revenue collection and public spending.

The sudden plunge in demand and increase in social expenditure following the collapse would have thrown member states’ budgets into disarray under any conditions, but the need to bail out domestic banks has added greatly to sovereign debt (Blyth 2013b). The institutional architecture behind the euro proved inadequate in preventing the rapid escalation of asset prices that underpinned much of the financial turbulence, and the EU only responded hesitantly and belatedly to the real threat of default created by the rapid deterioration of member states’ fiscal positions.

These crises fed into a third crisis: a demographic one. Many of the countries facing the most severe fiscal and economic crises – Italy, Spain, Greece, and Portugal – have among the lowest birth rates and the most expensive pension systems. At the same time, high unemployment, sluggish growth, and fiscal retrenchment have boosted Eurosceptic sentiment, which far-right parties across the EU are exploiting. The European Union’s migratory and demographic position exhibits tendencies that exacerbate the fiscal (and thus political) crisis. The countries most seriously affected by the crisis have the lowest birth rates (with Germany and other states in central and eastern Europe), while migrants, across northern Europe, and youth, across southern Europe, suffer from the highest unemployment rates. Because northern Europe does such a poor job of incorporating migrants into the labour market, and because youth unemployment in southern Europe is so high, immigration – a potential solution to the continent’s demographic problem – creates still further problems for Europe.

In all three areas, time, or more specifically timing, has not been on the EU’s side. The post-euro economic boom distorted peripheral countries’ view of their economic health and removed the incentive for substantial structural economic reform. Pressure for that reform only emerged through the crisis itself, meaning that reforms are being implemented at a time when fiscal retrenchment can only exacerbate a severe economic recession. Regarding institutions, the establishment of mechanisms at the EU level occurred only after the crisis was underway, when much of the damage was already done. The EU’s current institutional structure is a great improvement on what came before, but its gradual and reactive development meant that it failed to calm the markets rapidly and decisively, as an earlier action might have done. It is also ill placed to deal with sharp divergence in fiscal prospects that have emerged between northern and southern Europe. Finally, Europe’s demographic crisis has been well known to governments for decades, and yet the member states have only recently and tentatively spoken of migration as a solution to demographic decline. The challenge for the Union in the coming years is not only to find the correct solutions (itself a monumental task) but to implement them before they are obsolete. Put another way, the EU needs to find not only the right solutions but to find the right solutions at the right time.

Acknowledgements

We owe a great debt of thanks to David Howarth of the University of Luxembourg who provided extensive comments on an earlier draft of the article. Our thanks for comments also go to Mark Blyth, Pepper Culpepper, Peter A. Hall, and Amy Verdun.

Notes

1. In addition, in 2001, the European Commission warned Ireland about the dangers of excessive public spending (O’Leary 2010).

2. One might also say that the very manner in which the ECB targeted inflation was problematic: by targeting the rate of inflation in the euro zone as a whole, it would not be able to counteract country-specific trends, such as high inflation, that might cause problems down the road (Scharpf 2013: 116).

3. Some observers, such as Paul Krugman, have suggested even more ominously that demographic decline in Europe may be a recipe for ‘secular stagnation’. This is, however, still just a hypothesis.