The Policy Consensus Ruling European Political Economy: The Political Attractions of Discredited Economics


  • Waltraud Schelkle,

  • Anke Hassel



Since the Great Recession in 2008 academic economics has come under heavy criticism. But a straightforward alternative is not in sight either. We analyse in this article how the major flaws of applied economics are the mirror image of its attractions to policymakers, mainstream political parties and reform-minded administrations. We first assess what the consensus until recently has been and how it could have been implicated in the crisis. Secondly, we argue, following Hall, that the policy consensus continues to persist because it is politically attractive. The article ends with observations of how the management of the Euro area crisis still shows the attractiveness of the consensus.

Policy recommendations:

  • • Introduce greater pluralism in advisory bodies on economic policy to include civil society organizations, such as trade unions, and consumer organizations as well as academic disciplines other than economics.
  • • Review compensation policy for representatives on advisory committees to create a level playing field for experts from civil society organizations and independent consultants with financial sector lobbyists.
  • • Support and back up critical voices in the financial industry and in regulatory bodies against those who defend regulatory neglect and the privilege of rent-seeking for the financial industry.

1. Whose crisis?

Since 2008 crises in financial markets have forced governments to unprecedented monetary and fiscal intervention, unprecedented both in scale and degree of coordination. Sovereign debtors are under sustained attack by financial markets for their poor growth and pitiful public finances that the crisis of 2008–2009 generated. Governments are desperate for advice on how to stop contagion and a new recession. The situation is particularly difficult in Europe where the future of the European Union (EU) and the common currency in particular are at stake. This seems a pertinent moment for taking stock of what the economic policy consensus of the recent past has been, whether it is to blame for the recent crises or whether it can help policymakers now in their attempts at effective crisis management.

In similar previous situations, namely the Great Depression of the interwar years and the stagflation following the oil crises in the 1970s, the economic policy consensus of the time came under close scrutiny and eventually shifted to Keynesianism and monetarism, respectively. Today, criticism of academic economics is not in short supply but a straightforward alternative is not in sight either. The critics and mainstream economists do not agree even on how to label the consensus of recent years. The less favourable characterizations span from privatized Keynesianism (Crouch, 2009) and neoliberal market fundamentalism (Hall and Lamont, 2011) to macroeconomics based on models of a centrally planned economy (Buiter, 2009). For evidence of a paradigm in crisis, the critics can point to over-indebtedness of households and increasing inequality, but also to the unravelling of institutional fundamentals such as central bank independence.1

In defence, mainstream economics can point out that they explore market imperfections and outright failure in controlled variations of the perfect-competition-full-flexibility benchmark (for example, Smets and Wouters, 2003). The use of the consensus model by many central banks and supranational, applied research outfits proves that macroeconomists are not only scientists but also engineers engaged in fixing real world problems (Woodford, 2009). The defenders are bolstered by the fact that the lessons, which governments seem to have drawn from the various crises, suggest that the economic policy consensus of the last two decades is less obsolete than the critics think. As before, supranational policy reports and summit conclusions end with calls for structural reform, budget consolidation and commitment to price stability. Global financial regulation is still orchestrated by men in grey suits meeting in Basel and relies heavily on the same old instrument of (modest if somewhat raised) capital requirements. Put less favourably, a distinct possibility is that mainstream economics is a pathology (Hay, 2011) that simply lingers on without killing the patient outright.

But what does the consensus consist of? Those who see the financial crisis also as an ideational or intellectual crisis tend to characterize mainstream economics as the embodiment of neoliberalism that assumes that markets get it right. Those who acknowledge that there may have been some oversight but that there is no alternative to the new synthesis insist that there is a well-developed analysis of market failure in mainstream economics and it merely has to be updated in light of the new experience. An alternative to both, for which we argue in this article, is the position that accepts that it was not neoliberalism that got into crisis but a synthesis of neoclassical and new Keynesian economics that takes market failures into account; yet the alternative also concedes that the policy consensus has serious flaws that made it overlook all the factors leading up to this crisis, such as systemic risk.

2. New Keynesianism and neoliberalism – the argument in brief

Our contribution addresses the policy consensus on the European political economy in one fundamental way. In much of the writings on the shift from the post-war golden years to a new economic paradigm after the late 1970s, the contrast is drawn between the Keynesian welfare state on the one hand and neoliberalism on the other hand.2 Many political scientists and political economists therefore portray mainstream economics, its policy advice and the politics of market regulation as if it was under the spell of a decidedly neoliberal thrust (for example, Hay, 2011; McNamara, 2006; Stiglitz, 2008).

We think that this contrast is flawed. Keynesianism has not given way to a neoliberal agenda but to a new Keynesian-neoclassical synthesis that took price and wage rigidities into account, as a fact of economic life, with costs and benefits. The new Keynesian macro-economic policy that followed from it combined activist inflation targeting with structural supply-side policies, which means policies that aim at changing certain institutions like employment protection or wage-bargaining patterns.

Structural supply-side labour market policies are frequently seen as a defining element of neoliberalism. Activation, in the sense of privatizing the responsibility for finding a job, in contrast to an active labour market policy, was therefore seen as a cornerstone of a neoliberal agenda that abandoned the Keynesian welfare state. However, we argue that the turn to supply-side labour market policy was a response to political demands of core electoral constituencies rather than the biggest parcel in a neoliberal package.

In other words, both trends – the change in labour market policies and the demise of the Keynesian welfare state – went in parallel and were even connected, but not because of a macro-economic paradigm that was dismissive of the welfare state as such. At first glance, this difference sounds overly subtle, but we maintain that it helps us to understand the persistence of the policy consensus even after the financial crisis.

A challenge to our view is Crouch’s interesting hypothesis that the distinction between new Keynesian macro-economic policy and a supply-side policy agenda can explain why neoliberalism did not die (Crouch, 2009). Crouch’s answer is to claim that private Keynesianism succeeded the Keynesian welfare state, which kept demand steady and helped neoliberalism to continue even though it meant a harsher economic environment for the working population. He maintains that the Keynesian system of public demand management was not followed by a neoliberal turn to pure market rule, but rather to market liberalism combined with extensive consumer debt incurred by low-income and medium-income households (Crouch, 2009, p. 382). Privatized Keynesian demand management thus helped to maintain the rather unpopular and unwanted neoliberalism and liberalization.

We argue that new Keynesianism and supply-side policies interact in a somewhat different way. First of all, privatized Keynesianism, as portrayed by Crouch (2009), is not demand management at all but a reinforcement of the pro-cyclical movements of market demand. The new Keynesian policy consensus meant, above all, a move from the macro-level to the micro-level of economic management. This micro-level economic management addresses the supply side of the economy, that is, price and wage rigidities. It also tackles the denial of market access for certain consumers that is supposedly caused by a lack of competition between financial providers. However, liberalization cannot be sustained on a purely ideological basis for long. Policymakers must be seen to solve labour market problems and thereby respond to political demands by the electorate. We suggest that this is what the synthesis model delivered and why even a flawed model of new Keynesian policy-making became entrenched in our political system.

In the following, we will assess what the consensus until recently was and what its potential flaws were. Next, we explore what made the consensus attractive. In this, we follow Hall (1989, pp. 370–375), who proposed that ‘the political power of economic ideas’ requires, at a minimum, their economic, administrative and political viability. That is to say, ideas must resolve economic problems deemed pressing and relevant by policymakers; they must be in accord with bureaucratic practices and not overstretch implementation capacities; and they must appeal to broader constituencies and possibly allow policymakers to forge new coalitions. We conclude that, paradoxically, the flaws could be the flip-side of what made the consensus attractive, in particular to European governments with their perception of pressing low employment problems.

This is followed by an assessment how the economic policy consensus worked in practice and may have actually contributed to the crisis. The obsession with labour market reforms made policymakers miss the gathering storm in financial markets. The contribution ends with our observations of euro area crisis management that shows that the discredited economic consensus is still alive and well. The formation of politically cross-cutting coalitions that aim at defending the real economy against financial havoc has only started to form and the jury is still out on whether it will succeed.

3. What was the economic consensus of the last two decades?

This section presents the main elements of the workhorse model of mainstream macroeconomics, that is not its theoretically most advanced version but the analytical worldview with which applied economists are brought up. This workhorse model is called the new neoclassical synthesis or new Keynesianism; labels that can be used interchangeably, as we argue below.3 It is necessary to recall the basics because we want to revisit the critique from within that accuses the economic mainstream of a naive trust in markets and an obsession with general equilibrium in a complete market system (for example, Buiter, 2009; Krugman, 2009). These informed critiques are often taken up in more popular versions as the ‘neoliberal’ policy consensus in ideology and practice. In our view, the new mainstream has been much more interested in market imperfections and their policy implications than the critics acknowledge. This raises the question whether the economic policy consensus is at all to blame for the crisis. If it is to blame, we are in deeper trouble than even the critics think. It simply will then not do for economics to take account of the real world if that is what they already did.

The workhorse model for economic policy-making

The consensus model has three building blocks (Carlin and Soskice, 2006, pp. 81–90). In the first, aggregate demand (household consumption and possibly firm investment) are determined as resulting from income and the real interest rate – it is the conventional IS curve of the old neoclassical synthesis. In the second, the supply-side of the economy is characterized as resulting from wage and price setting in imperfectly competitive labour and commodity markets – this is the resurrected Phillips curve in a form that has absorbed the monetarist critique. This supply side determines a ‘natural’ rate of unemployment, given the structural and institutional features of the economy, such as transaction costs and corporatist arrangements that keep it from attaining full employment. In principle, this long-run equilibrium is compatible with any level of nominal prices. So we need, finally, a monetary rule that gives the economy a nominal anchor and gets it back into a low-inflation equilibrium after a shock. The central bank uses the interest rate, not money supply, which is the defining difference from monetarism. The rule describes the monetary authority’s preferences over the inflation–unemployment trade-off that characterizes the supply side of the economy.

How does a capitalist economy work in this stylized depiction? The standard situation is that the economy is in its long-run equilibrium and then hit by a shock, that is, an exogenous disturbance in demand (a change in investment or consumption) or in supply (a change in input prices), which pushes the economy off track. Since the ‘natural’ (un-) employment rate is determined by wage and/or price setters, this leaves only inflation as ‘a choice variable for policymakers’ (Akerlof et al., 1996, p. 1), typically with zero as the optimal inflation rate.4 The central bank perceives shocks as making the economy deviate from its inflation target. If then prices rise more, the central bank must raise the interest rate (or vice versa), which reduces the demand for credit that would sustain the existing level of investment and consumption. Higher unemployment will dampen wage and price increases, depending on the features of the labour and product market such as employment protection or costs of price adjustment that determine short-run trade-offs between inflation and employment. The central bank moves the economy along these short-run Phillips curves back to the long-run equilibrium. The more inertia there is in price and wage setting, the longer this will take and the more unemployment will be necessary to force down inflation. From the point of view of the central bank, the supply side (the Phillips curve) is thus the constraint on its stabilization policy while the demand side (the IS curve) is the transmission channel through which monetary policy works.

The first observation that will strike most readers is the central role of monetary policy for the working and stabilization of the economy. The first surveys of Goodfriend and King (1997) and Clarida et al. (1999) codified the macroeconomic consensus by pointing out the role of monetary policy in it. The consensus could be formulated without any reference to fiscal policy.5 The new synthesis considers fiscal policy to be distorting, determined by a political process and thus ruled by other than efficiency considerations (Goodfriend and King, 1997, pp. 237, 245, 280). In their extensive survey of the new Keynesian consensus on the conduct of monetary policy, Clarida et al (1999 p. 1702) mention fiscal policy only once, namely, when they note that in a low inflation environment, nominal interest rates may hit the zero bound and so the ‘important open’ question arises as to ‘whether cooperation from fiscal policy is necessary’ (p. 1702). It would probably have perplexed Keynes to find that this is considered an open question by economists who align themselves with his name.

Monetary policy here is ‘activist’ (Goodfriend and King 1997, p. 256), in the sense that the central bank does not simply wait and see after a disturbance. But the monetary authority is also not pro-actively seeking to shift the long-run equilibrium – this is the role left for government and their structural reforms of labour and product markets. The monetary rule is therefore a response function, summarizing (averse) preferences over inflation and unemployment. A central bank with high preferences for price stability chooses a radical disinflation strategy even if costs in terms of employment are high, and chooses a more gradualist one if it is less inflation-averse.

An activist central bank’s own preferences over the short-run Phillips trade-off that capitalist economies face imply a credibility problem. This arises when the central bank is more unemployment-averse than wage and price setters and can manipulate the very constraint it faces, here: inflation expectations. This difference in preferences seems to be a far-fetched assumption, given that central bankers are not normally recruited from the rank-and-file of trade unions. The rationale offered is that even independent central banks may come under pressure from governments with an inflation bias (Carlin and Soskice, 2006, p. 167). This has allowed intense study of how a central bank with an inflation bias can be committed to price stability, and was extremely influential in the design of the European monetary union (Blinder, 1997; Schelkle, 2006).

The consensus as synthesis

In what sense does this amount to a policy consensus and a new synthesis of once opposed schools of macroeconomic thought, Keynesianism and monetarism? In methodological terms, mainstream economists have come to accept most of the critique that neoclassical macroeconomists, from Milton Friedman to Robert Lucas, launched against the old neoclassical synthesis, meaning the pump-priming of the IS-LM model. Expectations must not be treated ad hoc but given careful thought, which has been turned into the stipulation that expectations must be rational in the sense of model-consistent.6 This has been generalized to a call for micro-foundations; that is, all macroeconomic relationships and responses must be grounded in individually rational behaviour that translates into aggregate behaviour. The synthesis makes extensive use of the eternally living representative agent that optimizes over an infinite time horizon. It is the basis for dynamic stochastic general equilibrium models that many central banks and the European Commission now use (Buiter, 2009).

In substantive theoretical terms, adherents of the synthesis accept the crucial role of monetary policy for stabilizing the economy and that even in the presence of under-employment, inflationary pressures may arise. It is a moot point which school of thought compromised more here. Monetarism and its offspring, real business cycle theory, claimed that the central bank should simply follow a strict money supply rule that endows the economy with enough additional transactions media to grow at price stability. This has been replaced by an inflation-targeting central bank that uses the interest rate actively to stabilize an economy prone to shocks. At the same time, (macro) economists have adopted the monetarist lens of the central bank that grasps the entire macro-economy by adding the reinterpreted Phillips curve as a constraint on its policymaking. The Phillips curve is now an aggregate supply curve fixing output and equilibrium employment in the long run and the adjustment path in the short run. It makes labour and product markets ultimately determine equilibrium while financial markets play only a residual role.7 One can thus argue that, strictly in terms of economic theory, the notion of a neoclassical synthesis is more pertinent while the policy activism of the central bank can be characterized as Keynesian. The terms can thus be used interchangeably, depending on whether the authors want to stress the supply-side determined, ‘natural’ equilibrium or the ‘realistic’ imperfections that can give the demand side and monetary policy a lasting influence.

The reconceptualization of the price mechanism is the most relevant example to illustrate what a rich research programme opened up thanks to this synthesis: the evidence from mature economies suggests that firms are not price takers but price setters in markets for less than fully substitutable goods or ‘brands’, that is, in monopolistically competitive markets (Akerlof et al., 1996, p. 21; Goodfriend and King, 1997, p. 249). The price setting that corresponds to this form of competition, namely mark-up pricing, can grasp a rich set of economic phenomena, such as pricing-to-market in volatile markets and smoothing over the business cycle. In contrast to marginal cost-pricing in atomistically competitive markets, it provides a surplus that can be the subject of negotiations with organized labour. Inflation can thus arise from ‘distributional conflict [among] different social groups’ held only in check by a credibly inflation-averse central bank (Carlin and Soskice, 2006, pp. 133–134, 160–168).

Its flaws in general

Our overview of the economic policy consensus before the crisis contained three possible candidates that may be relevant to the question ‘Whose crisis – that of neoliberalism or of something called new neoclassical synthesis/new Keynesianism?’ It may be helpful to summarize them briefly at this point, mainly to show that what could be dismissed as rather esoteric squabbles among academics before the crisis might be serious flaws contributing to the crisis from hindsight.

At the most obvious level, we agree with the critics of mainstream economics that the underlying benchmark of a dynamic general equilibrium is a problem (see Prosser [2006] for a particularly insightful critique). This reference point gives the impression that the enlightened Visible Hand can shift the economy gradually and continuously towards this benchmark by getting rid of rigidities, by aligning incentives through more transparent information and by allowing for the emergence of missing markets through permissive regulation. Financial markets cannot alter this underlying equilibrium; if anything they should facilitate its attainment. This is the source of the label, ‘market fundamentalism’ for the trend in economic policy over recent decades (Hall and Lamont, 2011). It also paints a much more optimistic image of capitalist economies than we can find in Keynes (1936). In this mainstream economics perspective, capitalist economies may be full of microeconomic imperfections but they have no systemic flaws. Market adjustment may work imperfectly but it does not work perversely, as Keynes (1936, p. 291) maintained for a situation of deflation, to be outlined presently, and Shiller (2003) claimed for a situation of asset market bubbles.

Secondly, the micro-foundations agenda, while seemingly esoteric to non-economists, served to restore the superior role of price adjustment. Only non-economic explanations, such as political forces (insider power) or psychology (the representative worker resents inequality), can make sense of ‘real rigidity’, for instance wage earners resisting wage cuts even when faced with rising unemployment. This goes directly against the Keynesian proposition that quantity adjustment may trump price adjustment in capitalist economies and that this may be functional. For the latter, take the case of why market forces cannot lead an economy out of deflation: even if unemployment has risen, there may be no tendency for real wages to fall. Rising unemployment makes desperate workers to offer their services at ever lower wage rates while desperate firms lower prices to sell the goods they already produced. But if both wages and prices fall, then real wages stay roughly constant and in any case do not necessarily fall as much as needed to make firms keep their work force, let alone hire more workers at lower nominal wages. Thus, the synthesis does not consider the case that nominal flexibility can generate real rigidity, and arguably suffers from a fallacy of composition. It takes the whole for its parts, here: by assuming that both wages and prices are rigid while they can be both flexible, making the whole rigid.8 In macroeconomics, (nominal) flexibility can be the problem rather than the solution. But the micro-foundations agenda is based on the premise that macroeconomics is not a field of study in its own right because the whole is equal to the sum of its parts, neither more nor less. This is too strong an assumption, as Keynes (1936, pp. 358–361) illustrated also with his ‘paradox of thrift’9 and social choice theorists like Thomas Schelling (1978) have confirmed since then.

Finally, the consensus model has a sensible take on the Phillips curve, namely that it ‘may exist but it cannot be exploited’ (Carlin and Soskice, 2006, p. 75). Yet, by making it the prime constraint on monetary policy and macroeconomic stabilization generally, it focuses all attention on labour and commodity markets. Financial markets come in only as an afterthought regarding the transmission of monetary policy. Policymakers were busy looking under the lamppost of the new synthesis for yet another necessary structural reform while financial bubbles were allowed to grow. This was despite the fact that there have been plenty of warnings about asset market bubbles (even Alan Greenspan admitted that there may be irrational exuberance in financial markets), yet the moment passed after markets rebounded after the 2001 crash. By concentrating on markets for (the flow of) services and goods, there was an in-built analytical bias against considering the catastrophic stock flow dynamics resulting from asset and debt accumulation. These dynamics would come to trump any rigidities in labour and commodity markets upon which inflation targeters and structural reformers so obsessively concentrated.

4. What were the attractions of this economic policy consensus?

The flaws of the new synthesis did certainly not diminish its political attractions. They may even have contributed to its attraction, which we can understand following Hall (1989) by assessing its economic, administrative and political viability. He synthesizes three approaches. There is, firstly, an economist-centred account that claims it is expert advice in government that gives economic ideas powerful influence; which is typically proposed by academics who served for some time in government or in a central bank (recent examples can be found in Bussière and Stracca, 2010). Second, a state-centred account claims that the extent to which economic ideas catch on with the bureaucracy and, in particula,r senior officials is crucial for their success, an approach initiated by Theda Skocpol in historical-institutionalist studies (Weir and Skocpol, 1985). Finally, a coalition-centred account of policymaking, in the version that Peter Gourevitch (1986) championed, stresses that the brightest economic ideas do not have much effect if there are no coalitions of interests organizing around them. Hall (1989, pp. 8–13, 370–388) suggests that each of these accounts notes an important requirement for the viability of an economic idea in practice: it must be translated from a scholarly discourse into something that policymakers find useful; the state machinery must be able to work with this idea; and a set of political constituencies must find policies based on the idea to be in their interest.

Obviously, each of these determinants and their combination leave a lot of space for country variation that would require a research project of its own to fully explore here. We illustrate the argument using the case of Germany. The German case is crucial because it shows how strong the pull towards supply side in employment policies had become. The so-called Hartz reforms in the early 2000s were a clear departure from earlier policies that were characterized by the ‘welfare without work’ syndrome (Esping-Anderson, 1996). German unification had led to an expansion of social policies that aimed at making unemployment bearable rather than forcing potential labour supply into the market. This changed, notably under a government led by Social Democrats, and since then long-term unemployment has given way to in-work poverty. While quite contested at the time, we show how this shift in labour market policy was supported by a decisive political consensus that supports Hall’s conjecture.

Economic viability

The consensus model must have offered answers to what representative politicians or expert audiences outside central banks perceive as the most urgent economic policy problems. For this to be the case, they should not be required to understand the workhorse model in any detail. During the 1980s inflation had been brought under control by central banks mandated to focus on price stability but the rising levels of unemployment with which economies entered every new business cycle remained a pressing concern. So, first of all, a model that moved from monetarist inflation-fighting to responsive inflation-targeting was welcome, thereby conceding that heavy-handed inflation fighting has had high costs in terms of unemployment. Moreover, the model took into account all those factors that could explain the ratchet effect in unemployment levels. Market failure, institutional rigidities and hysteresis effects like the rapid devaluation of human capital were all enlisted to explain rising levels of equilibrium unemployment.

The downward rigidity of nominal wages was not a problem in an era of moderate and variable inflation. In fact, downward rigid money wages were helpful for real adjustment because changes in the price level or the exchange rate could then engineer changes in real wages across the board, leaving the wage structure of different types of workers relatively untouched. But after the breakdown of the Bretton Woods exchange rate system, inflation (expectations) became much less controllable as the nominal anchor of a dollar standard had been removed. It took the volatile 1970s to build up the resolve among policymakers to fight inflation head on, starting with the Volcker shock in 1979. For the Europeans, this also meant stabilizing exchange rates and this started monetary policy coordination that led to monetary union in 1999. These ultimately successful attempts at lowering and stabilizing inflation had the effect that downward rigid nominal wages became rigid real wages; when the economy went through a phase of disinflation, real wages could even rise as a result. Coordinated wage bargains, which had not only introduced wage floors but also wage ceilings and used to standardize wages across industries to pre-empt the poaching of skilled workers, became dysfunctional even though they had on the whole a levelling effect on wage growth.

Organized labour defending wage coordination was then accused of serving insiders only, to the detriment of the unemployed, female and young entrants into the labour market. This was an accusation that especially the OECD jobs study of 1994 popularized and at the same time managed to give it the air of rigorous economic analysis. Some relief came from work that showed that not all wage coordination was bad. Collective wage setting can keep wage increases at a competitive rate if they internalize the possible damages of overly generous settlements for the economy as a whole. Such beneficial coordination can come from monopoly unions or from strong unions in the exposed traded goods sectors that set the ceiling for all others (Calmfors and Driffill, 1988; Soskice, 1990). But where these institutions or such an export-orientation of a national political economy are absent, labour market flexibilization was the only game in town.

A late manifestation of the new consensual approach was the European Council’s endorsement of ‘flexicurity’ in December 2007. In the words of the Commission, it ‘involves the deliberate combination of flexible and reliable contractual arrangements, comprehensive lifelong learning strategies, effective active labour market policies, and modern, adequate and sustainable social protection systems.’ (European Council, 2008, p. 52) The flexicurity concept had the beauty of being considerate as regards the diversity and the complexity of social policy and labour market interactions. As the recent report on employment in Europe points out:

Everything considered, there is no single combination of policies and institutions to achieve and maintain good socio-economic results, but rather there are different pathways to good performance that are, to a large extent, the result of distinct historical trajectories. Respecting the principles of subsidiarity (and the Open Method of Coordination), this allows scope for tailor-made policy packages to suit national preferences with respect to distributional aspects, risk-taking and other national objectives (European Commission, 2008, p.177).

This is perfectly in line with the consensus: supply-side reforms can shift economies towards a more beneficial fundamental equilibrium and increase employment; this holds notwithstanding institutional diversity.

Administrative viability

The record on administrative viability is arguably more mixed. On the one hand, the demands on administrative capabilities in macroeconomic policy diminished to the extent that responsibilities for stabilizing demand management moved to central banks. This has enormous practical advantages. It requires only a meeting of the central bank’s governing council or monetary policy committee to change the interest rate and possibly other conditions under which banks may refinance their credits to the private sector. By contrast, discretionary spending programmes are full of practical pitfalls for the executive. They give the opposition in parliament an opportunity to accuse the government of ‘too little, too late, or any other easy criticism that ad hoc programmes deserve. They require bureaucrats to find temporary jobs in sufficient quantity but without too much crowding out of existing private sector capacities. They ask for income to be put at the disposal of those who spend it (instead of saving it) without too much leakage or fraud. The lags in fiscal policy literature initiated by Friedman (1953), summarizes these difficulties in a number of timing problems. The synthesis model that elevates the central bank to the prime stabilizer of the macro-economy and leaves fiscal policy to rely on automatic stabilizers thus came as a great relief.10

However, administrative capacities became stretched in other respects, namely by the microeconomic (‘structural’) supply-side reforms that governments were instead meant to engage in. They could all be justified as moving the long-run Phillips curve towards lower equilibrium unemployment. But it is a complex task to operate ‘activating’ labour market policies, such as putting recipients on training programmes or engaging in individual case work for job placement; using the tax system for employment-friendly subsidies and rate structures or writing contracts for private providers of welfare services that are closer to markets but also have market incentives. This called for a profound reorganization of bureaucracies, for instance, amalgamating public employment services and welfare offices to ‘one-stop-shops’ where unemployed beneficiaries can get the full range of offers but benefits can also be used as a sanction to monitor the effort in job search (Schelkle et al., 2012). These complexities were tackled by new public management techniques and by engaging private providers for frontline work.

The EU played a facilitating role in this. As governments were picking up the trend and went for activating labour market policies, which included outsourcing to private providers, the EU helped to build networks of ‘enlightened’ civil servants who understood that they could not simply oppose this trend. An example of how successful this can be is the formation of the active European network of heads of public employment services (HoPES). To this very day, HoPES embraces and shapes the activation and outsourcing reforms that governments want so as to keep pivotal role for the public sector rather than succumb to retrenchment (Weishaupt, 2010).

The German Hartz reforms are a good case in point. As described in detail in Hassel and Schiller (2010), activation policies in Germany originated in the various layers of public administration among bureaucrats. In informal meetings of leading civil servants the ineffectiveness of the existing system was raised and reform proposals contemplated (Fleckenstein, 2008; Hassel and Schiller, 2010, p. 212ff). The most pressing urge for change came from local authorities, which were weighed down by increasing numbers of long-term unemployed people with no prospects or incentives to find employment. The system was perceived as wasteful and counterproductive; the labour agency had a reputation of being a self-satisfied bureaucracy. Local-level activation pilot projects were carried out in several places, often with the help of private think tanks, in order to prove that the practice could be changed. The promoters of new activating approaches were then integrated in national reform policies when the government changed.

The real push for structural reforms came from the looming fiscal crisis. As the dotcom bubble burst and the German budget violated the deficit limit of the Stability and Growth Pact in 2000, the government began actively and seriously to look into structural reforms to curb social spending (Hassel and Schiller, 2010). Activation through restructuring unemployment benefit entitlements was the solution that combined the urge for creating job incentives for low-skilled people with controlling social expenditure.

Still, the reorientation from macro to micro management has proved to be an arduous, often costly process and is by no means resolved. But we can see that the micro-foundations turn had attractions for fiscal authorities at the time, since structural reforms promised to let the state off the hook with respect to difficult-to-implement stabilization programmes that can quite visibly fail.11 At the same time, this turn gave the state a role in modern social engineering, supporting an active bureaucracy rather than the complete retreat of the state.

Political viability

Regardless of its administrative viability, the supply-side agenda contributed to the political viability of the consensus. The new synthesis appealed in particular to centrist social democrats who had struggled with the weakening of their electoral base of organized labour in manufacturing for some time (Kitschelt, 2000; Pontusson, 1995). Centre-left policymakers not only faced the problem of a steady decrease of industrial employment rates but also increasing political pressures from the trend in inequality. Their core constituencies became divided over the amount of social spending that went into transfer payments for labour market outsiders. Social democratic governments were criticized for collusion with insiders, supported by (spurious) evidence that they were less likely to increase spending on active labour market policies benefitting outsiders (Rueda, 2006).

In electoral systems of proportional representation, centre-left parties in the 1990s frequently faced the dilemma of either protecting the interests of the lower middle-class and dominating the centre-left but not winning an electoral majority, or moving towards the median voter and losing the support of traditional supporters (Kitschelt, 1999). A successful electoral strategy implied the move towards the centre ground.

Supply-side reforms were thus a welcome opportunity, firstly, to shed the image of social democrats being hooked on pump-priming and redistribution, and, secondly, to conspicuously do what is economically sensible even if it hurt their own constituencies of labour market insiders in terms of job security (while promising them a larger take-home pay packet in the long run). Structural reforms aimed at securing the centre ground of electoral competition by orienting centre-left parties towards the median voter.

We can thus see how labour market reforms became attractive for parties in the section of the ideological spectrum where one would least expect it. The divisions over preferences of social policy reforms enabled governments to engage in structural reform that changed the distributional effects of employment-related social policies (Häusermann, 2010). Centre-left parties hoped to claim credit for activation policies and family-friendly social investment strategies from female carers and workers as well as other labour market outsiders (Bonoli, 2012).

From the early 1990s onwards, labour market activation was indeed promoted primarily by centre-left parties and executed by centre-left governments. Examples for a turn towards activating policies by centre-left governments are Denmark, the UK, Sweden and Germany (Bonoli, 2010). Under the conditions of budgetary constraints, activation was an alternative to straightforward retrenchment by the centre-right and were still in line with fiscal retrenchment.

Germany, as the latest example is again a case in point. Within the Social Democratic Party (SPD) a lively discussion about the electoral strategy had erupted from the early 1990s onwards. The election campaign in 1990 led by traditionalist Oskar Lafontaine left the party with the lowest result in post-war history, namely only 33 per cent of the popular vote. A leadership battle in the mid-1990s led to a shift to the centre-right under the chairmanship of Gerhard Schröder who pushed the SPD vote to over 40 per cent in 1998. After the change of government, leadership positions within the parliamentary group and the Ministry of Labour were filled with ‘modernizers’ in social and labour market policies (Hassel and Schiller 2010, p. 140ff).12

The party leadership hoped that through a decisive move towards structural reform policies the ongoing internal conflicts would be solved to the advantage of the party modernizers. A modern centre-left would create new coalition possibilities as the Social Democrats’ embrace of structural reforms was quite compatible with the manifesto of Liberals and even the compassionate wing of Conservative parties. Thereby a broad party-political spectrum could be rallied around the structural reform agenda that the new synthesis supported.

Moreover, the comparative policy evaluations that supranational agencies like the OECD and the EU produced relentlessly spread the message that activation is best practice, successfully operated in impeccably social democratic countries such as The Netherlands. Finally, structural reforms also opened up room for projects of local job placement, which were popular with politicians due to their experimental character and controllable effort, and the opportunity to claim responsibility for them.

Treasuries, but also senior officials in spending ministries, endorsed such public sector modernization. The hallmark of this economic policy consensus was the Lisbon Strategy decided in 2000, when a majority of centrist social democratic governments had just come to power in the EU. The officially ordered critical reviews of this Lisbon Agenda, most notably gathered together in the Sapir et al. (2004) and the Kok (2004) reports, criticized less the substance of the consensus than the lack of resolve with which governments pursued it.

5. Conclusions: the continuing political attractions of a discredited consensus

We have shown that there was clearly a consensus on macroeconomic policy that was not monetarist or of the real business cycle variety – or what political scientists would call neoliberalism. Hence, we do not share the critique that part of the problem that got the rich world into the worst recession in the post-war era was that economists were hooked on models irrelevant to an imperfect world and supported a neoliberal market fundamentalism that wanted to get government out of the way. The consensus research programme consisted of an intense analysis of market imperfections and it endorsed active policy interventions, especially on the supply side. This consensus was not purely academic: it also reached into the research departments of the central banks, the International Monetary Fund, the OECD and the European Commission, which ultimately informed the building of econometric models in policy analysis and evaluation.

However, we agree with the critics that the reference to a benchmark of a general dynamic equilibrium was a source for misleading policy advice. However, in our view it is not so much the assumptions of perfection that were problematic, since they were systematically scrutinized in subsequent analyses. With hindsight, we can see that the assumption of it being fundamental was the problem, since this is the economist’s jargon for ‘determined in the real economy’. This analytical anchor led to a blind spot towards the role of financial markets and monetary policy in determining the activity level of any economy.

Two other flaws are not recognized as fully as we think is necessary. There is, firstly, the preoccupation with micro-foundations, that is, the idea that all macroeconomic phenomena must be derived from individual optimization, which made macroeconomists neglect systemic risks and rendered them susceptible to fallacies of composition. Secondly, the focus on (imperfect) labour and commodity markets left no specific and fundamental role for the financial system and therefore missed how the malfunctioning of asset markets may feed back into commodity markets.

Unfortunately, we also found reasons to believe that it was exactly these flaws that contributed to the attractions of the economic policy consensus. Policymakers and administrations were attracted to micro-optimization in the context of a fundamental equilibrium model because it deflected attention from governments’ perceived weakness in macro-steering and offered plenty of policy choices at the micro-level. This came in handy as the composition of the electorate changed, rapidly leading to a search for the new middle but also to cross-party coalitions. The consensus also seemed to address the most urgent economic policy problem of our times, namely, how to raise activity rates without pump-priming the economy into inflationary growth. It was respectful of diversity and the functional equivalence of, say, individual wage contracting and coordinated wage setting in achieving good economic results. Hence, it did not require the elimination of all institutional diversity, which made it particularly popular with governments and the Commission in the EU.

The political demand for the policy consensus was therefore strong and continues to be so even after the financial crisis, if the reforms of economic governance in the Economic and Monetary Union since 2010 tell us anything about revealed preferences of governments (Gros et al., 2010). The policy debate about what next? on both sides of the Atlantic is hooked on the need for fiscal restraint, on the one hand, and stimulating private consumption and investment by ending the credit squeeze, on the other. This is still perfectly in line with the pre-crisis consensus in its emphasis on a secondary role for fiscal policy (that should be restored) and on creating demand through private credit expansion.13

This step back to the future is quite worrying, however, if the economic policy consensus of the last two decades is partly to blame for the financial crisis. In their first reactions, governments in the G20 and leading member states of the EU seemed to be determined to end the regime of self-regulation in financial markets (Sarkozy, 2008), make the regulatory regime less dependent on private credit ratings (Yassin, 2010) and generally rein in excesses like scandalous banker bonuses that are paid out regardless of the performance of the businesses they manage. While some robust re-regulation of financial markets is in the making, there has also been some back-tracking, for instance by making the new emergency facility of the European monetary union dependent on private bond finance and hence credit rating. There has certainly been no end to scandalous bonuses paid by the very same banks that were bailed out by taxpayers who still suffer from the consequences of the crisis.

There is some alternative thinking in economic theory available, for instance at the Institute for New Economic Thinking (n.d.) sponsored by George Soros. But the alternatives were pushed into heterodoxy and had scholars had little chance to put forward robust models for policy analysis, in exchange with professional users in central banks and treasuries. The herding behaviour concerned not only financial markets but economics in academia as well. (Buiter, 2009; see also Davis, 2011. Thus, we are not optimistic that change will result merely from changing our economic models. But the cross-cutting coalitions that could form around the aim of defending the real economy against financial havoc have yet to come forward and supersede the political clout of the financial industry.

What could move things on from here? One policy conclusion is certainly that more pluralism in institutionalized policy advice would help. Councils of economic advisors or committees on reform should have a wider catchment area than the economics departments of renowned universities and the upper echelons of business. Representatives of trade unions and consumer organizations but also from other academic disciplines should have guaranteed places. The quality of advice does not merely consist of understanding estimates and models put out quickly to silence entrenched interests but of preparing decision-makers for the diversity of concerns and distributive effects that may hit them once a policy proposal is put into the public domain.

This will require some innovation in recruiting advisory committees. For instance, the cognitive capture by the financial industry is perpetuated not least by the fact that financial institutions can afford to pay their representatives on the various supranational committees while independent consultants, consumer organizations and organized labour cannot afford this. The policy to ensure independent advice by not paying those experts and stakeholders that sit on a myriad of EU committees thus leads to their giving inferior advice. This seems to have been recognized by the Internal Market Commissioner Barnier who is in the forefront of changing the composition and the compensation of committee membership in financial regulation (Austrian Federal Chamber of Labour, 2011).

Last but not least, policymakers and governments can strengthen those voices within the financial industry and the financial media that articulate self-criticism and endorse moves that would protect the financial industry from itself. Analogous to big business at the turn of the last century, which was crucial for the emergence of employment protection and social insurance (Swenson, 2002), those members of the financial elite who show some insight into past failings and flexibility of mind must be coopted and put in charge of reforms. Such brave and insightful characters do exist. But they also need the courage of political office holders to stand by them when those who defend regulatory neglect and the privilege of rent-seeking for the financial industry try to discredit the reputation of those who they see as traitors. The recent stand-off between Barclay’s, represented by Bob Diamond, and the Bank of England, represented by Paul Tucker, over the fixing of the Libor rate is a prime example for the challenges ahead. It seems that the Bank of England has just about won in this stand-off. But it will require many more such small victories in order to undo the legacy of a discredited policy consensus.


We thank participants at the Dahrendorf symposium in November 2011 and at the European Institute’s student–staff seminar at LSE for helpful discussions and are particularly grateful for constructive comments from Abel Bojar (LSE), Eddie Gerba (Kent) and Deborah Mabbett (Birkbeck). The usual disclaimer applies.


  • 1

    See Hodson and Mabbett (2009) for an insightful analysis of the UK in 2008–2009 and Schelkle (2012) for the European Central Bank.

  • 2

    See, for instance Jessop (2010) for the contrast between the Keynesian welfare state and the Schumpeterian workfare state but also Hall (forthcoming): ‘Active labor market policies were the supply-side alternative to Keynesianism. Although their complexion varies from one country to another, they involve government subsidies for training positions or jobs created for groups at greatest risk of falling out of the labor market, such as the young and long-term unemployed’.

  • 3

    See Goodfriend and King (1997, sect. 5) and Clarida et al. (1999, p. 1662) for the first overview articles that noticed the consensus among academic economists and practitioners of central banking; for more recent and accessible presentations, see Carlin and Soskice (2006: especially ch. 15) and Woodford (2009). Mankiw and Romer (1991) have presented an early (new Keynesian) textbook version.

  • 4

    In his comments on Goodfriend and King (1997, p. 249), Blanchard (1997, p. 293) notes that the presence of wage rigidities makes the consensus that central banks should target inflation close to zero doubtful.

  • 5

    The reverse also holds: Carlin and Soskice (2006, ch. 6) devote an entire interesting chapter to fiscal policy but do not use their workhorse model at all. In other words, the model is not useful for the analysis of fiscal policy.

  • 6

    That is, economic agents modelled exploit all the information available (which does not have to be complete or perfect) and this information includes the model itself. This methodological principle was first stated by a mainstream Keynesian economist, John Muth (1961).

  • 7

    See Akerlof et al. (1996) and Galbraith (1997) for a critique of the old and new synthesis. Keynes’ economics arguably had the employment level determined in the interaction between product markets (effective demand, in the sense of demand expected by firms contemplating investment) and financial markets (taxing the real economy with its demand for an interest rate that is determined not by the readiness to defer consumption but by the readiness to give up liquidity).

  • 8

    Economists hooked on micro-foundations typically take a short-cut and pretend that workers and firms directly contract a real wage because the actual price level is equal to the rationally expected level. See Carlin and Soskice (2006, pp. 46–47) who at least problematize this short-cut.

  • 9

    The paradox is that households’ attempt to save more can lead to less saving in the economy. It is based not on some rigidity but on the perfect functioning of the price mechanism that responds to lower demand when goods are actually bought and sold for money payments.

  • 10

    Chapter 6 on fiscal policy in Carlin and Soskice (2006) starts not with discretionary spending policies but automatic stabilizers, that is, built-in revenue and expenditure items of a budget that vary with the business cycle such that the balance moves counter-cyclically.

  • 11

    The composition of stimulus packages in OECD countries during the crisis of 2008–2009 provides evidence for the fact that governments have not gone back to large-scale public employment programmes (IMF, 2009, table 5).

  • 12

    None of the ‘traditionalist’ members of parliaments achieved high office in the field of social and labour market policy after 1998. The Minister of Labour was a politically inexperienced and modernist trade union leader who initiated a far-reaching and predominantly liberalizing pension reform (Hassel and Schiller, 2010, pp. 184ff.).

  • 13

    In this vein, Carlin and Soskice (2006, p. 571) discuss consumption smoothing of households as resulting purely from the counter-cyclical credit demand of rational households to anticipated interest rate policy.

Author Information

Waltraud Schelkle is Senior Lecturer in Political Economy at the European Institute of the London School of Economics and Adjunct Professor of Economics at the Free University Berlin.

Anke Hassel is Professor of Public Policy at the Hertie School of Governance. She is also an Adjunct Professor of the Graduate School of Social Sciences at Bremen University.