The IMF’s new conditionality

Crafting change, lessons from Eastern Europe

18 February 2011 | At Issue

A fully-formatted, but unreferenced PDF version of this briefing is also available.

The 2007-2010 crisis offered the International Monetary Fund an unexpected opportunity to demonstrate that it was serious about changing its emergency lending practices, heavily criticised for being unnecessarily contractionary and drawing on outdated economics (Easterly, 2004). The new conditionality policies promised increased flexibility, minimal structural conditions and mainstreaming inflation targeting as the new economics of crisis (Gabor, 2010a).

An increasingly popular monetary policy framework, inflation targeting works as follows: the central bank establishes (together with the government) a target for inflation that is consistent with the economy producing at full potential. The central bank commits to achieve this inflation target over a time horizon (typically one to two years) by manipulating its interest rate, to the exclusion of achieving other targets such as level of employment or an exchange rate. The theoretical framework informing inflation targeting regimes offered the IMF an avenue for reforming the economics of its conditionality (known as financial programming) without altering its underlying principles:

  1. Economic stability is best achieved through price stability (rather than employment or growth), a view also promoted by financial programming. The new policy framework envisages that price stability can and must be pursued through the central bank’s influence over aggregate demand: changes in interest rates bring the economy back to its potential and inflation to target.
  2. Economic policy, particularly during crisis times, can only be formulated ‘rationally’ if insulated from political interventions (Woods, 2006), in other words assigned to politically independent central banks. In fact, inflation targeting supporters contrasted the sophisticated monetary ‘science’ of modelling dynamic individual behaviour with the ‘alchemy’ of inevitably politicised fiscal decisions (Leeper, 2010). Thus, by embracing inflation targeting models, the IMF could claim “scientific foundations” for its (criticized) insistence that low inflation should be the policy priority.
  3. Fiscal policy not only lacks theoretical foundations but is ineffective: rational agents recognise future tax burdens created by fiscal activism, and reduce present consumption/investment accordingly.
  4. Markets, and not central banks, are best placed to set exchange rates. Neither concerns with competitiveness nor precautionary accumulation of reserves are necessary as monetary policy credibility ensures access to international financial markets.

Even before the crisis, critical voices questioned the wisdom of instituting a policy more concerned with inflation than unemployment as the framework for addressing crises, particularly since it suited the financial sector rather than development agendas (Epstein, 2006). The IMF dismissed such concerns, instead providing countries with technical assistance to help speed up the transition. However, it had limited opportunities to test how conditionality might work under inflation targeting (only in Colombia, Brazil and Peru), because emergency lending was concentrated in low-income countries yet to adopt this policy regime. The collapse of Lehman Brothers in September 2008 changed this, bringing crisis to Eastern Europe.

Eastern Europe problems

Eastern Europe’s impressive growth performance pre-crisis relied on a consumption-driven model enabled by short-term capital inflows and appreciating exchange rates (Fabrizio et al, 2009). As in high-income countries, the prevailing model of banking sharpened vulnerabilities. The foreign-owned banks dominating banking sectors (an outcome of post-socialist privatisations) would borrow short-term in international money markets or from Western European parent banks to lend long-term in foreign currencies, resulting in housing and consumption booms and growing current account imbalances. By September 2008, over 50 per cent of housing loans were denominated in foreign currencies (euro or Swiss francs) in most countries in the region. Consumers accepted the exchange rate risk for two reasons: (relatively) lower interest rates and expectations that their national currencies would continue to appreciate (similar to expectations that US housing prices would continue to rise). However, the post-Lehman convulsions in international money markets saw regional currencies quickly depreciating and suddenly raised the possibility that parent banks might refuse to roll over the short-term debt of their Eastern European subsidiaries (Gabor, 2010b). Faced with such a ‘sudden stop’ in capital inflows, three European Union members, two with inflation targeting regimes already in place (Hungary and Romania), turned to IMF emergency assistance.

At a first glance, what is striking about the IMF’s conditionality is the small difference between targets designed under the traditional financial programming (for Latvia, given its fixed exchange rate regime) and those for inflation targeters. As Table 1 shows, an inflation band replaced the financial programming criterion for central bank credit (and only as an indicative target for Romania). In fact, the design of monetary conditionality conformed to traditional IMF programming, giving the impression that fiscal misbehaviour rather that private banking practices had forced countries to seek the IMF’s help; fiscal targets dominated conditionality.

Table 1: Selected indicators for IMF programme countries in Eastern Europe

(date of agreement)
(December 2008)
Romania (inflation targeting)
(April 2009)
Hungary (inflation targeting)
(November 2008)
Quantitative performance criteria A floor on the change in net international reserves, non accumulation of external debt arrears
  • Ceiling on net domestic assets of the Bank of Latvia
  • Floor on adjusted cash fiscal balance
  • Floor on adjusted primary cash fiscal balance
  • A ceiling on general government domestic arrears.
  • A floor on the overall general government cash balance
  • A floor on the central government system primary cash balance.
  • A band around the 12-month rate of inflation of consumer prices (IT)
Ceiling on public guarantees
Quantitative indicative target Ceiling on the general government wage bill General government primary spending (excluding EU funds) Ceiling on the total debt stock of the central government system.
A consultation band around the 12-month rate of CPI inflation (IT)
Structural benchmarks Pension and banking laws reform. Fiscal reforms (wage and employment), further privatisation of public services, increased taxation (VAT and excises) etc.
IMF predictions vs. actual GDP contraction 2009: -5% vs. -18%
2010: -3% vs. 0.5% (f)
2009: -4.1% vs. – 7.5%
2010: 0% vs. -1.9 (f)
2009: -1% vs.- 6.3%
2010: -0.4% vs. 0.1% (f)
IMF predictions vs. actual 2009 fiscal deficit
(% of GDP)
4.9% vs. 7%
(net fiscal tightening of 14% of GDP over 2009-2010)
4.6% vs. 7.3%
(real spending cuts of 6% of GDP over 2009-2010)
3.9% vs. 3.9%
(real spending cuts of 8% of GDP over 2009- 2010)

Source: adapted from (Gabor, 2010a)

What role for central banks?

In this context, it becomes important to examine the benefits of the IMF’s decision to overhaul its conditionality. First, inflation targeting models do not have a well-developed theory of fiscal policy (Leeper, 2010), so the IMF focus on fiscal conditions conflicts with promises to address the theoretical shortcomings of financial programming. Instead, country programmes experienced the typically messy processes of implementing spending cuts, aggravated by the IMF’s tendency to underestimate output contractions (see Table 1). Throughout 2009, both Romanian and Latvian governments were forced to renegotiate deficit targets as compliance with the original targets would have required even harsher spending cuts or higher taxes. The increased leverage allowed the IMF to demand additional structural reforms such as changes to pensions, contradicting its promise to reduce the structural component of conditionality.

Furthermore, to achieve targets for 2010 governments turned to drastic public employment and wage cuts (at 20 per cent and above in Romania and Latvia), with planned cuts to pensions eventually declared unconstitutional in Romania and Latvia. The IMF also advised tax reforms (higher VAT rates and excises) that sought to shift the burden of taxation to consumption, further depressing private demand and private job creation. The social impact will be familiar to voices critical of the IMF’s past conditionality: along with large recessions, unemployment doubled in Romania to around 8 per cent (underestimated by drop-outs from the labour force) and tripled in Latvia to 21 per cent in 2010. Sluggish growth, driven by a slowly recovering export sector, will see the region growing at half the speed of East Asia or Africa for the third year since the start of the crisis.

Second, and crucially, the usefulness of inflation targeting to guide policy decisions rests on the central bank’s ability to manipulate aggregate demand in order to bring the economy back to its potential and thus inflation on target. Nevertheless, the IMF programmes in Eastern Europe embed an apparent paradox. Despite the primacy given to monetary policy in the new economics of crisis and in contrast to the plethora of fiscal constraints, policy documents refuse to contemplate the possibility that central banks could play an active role in accelerating the recovery. The costs of this neglect, the IMF admitted, would be considerable: a negative impact on long-run growth and a delayed return to the lower growth potential. On the upside, according to the IMF, subdued demand would offset inflationary pressures arising from increased taxes and higher global food prices. Instead, IMF reports identified the following key constraint to central bank responses: “easing of interest rates and reserve requirements would be permitted, but only to the extent that exchange rate and the NFA [reserves] target permit” (IMF, 2009:18).

It is no coincidence that central banks’ interest rates in both Romania and Hungary were the highest in the European Union throughout 2009 and 2010, at levels usually found in fast-growing economies. Whereas previously the IMF would advise exchange rate devaluations to address balance of payments difficulties, during this crisis it insisted that interest rates decisions be tailored to arresting exchange rate depreciations in order to prevent the deterioration of banks’ balance sheets. This is important because it shows how heavily politicised the IMF’s position towards exchange rate interventions under inflation targeting is. It demands exchange rate flexibility when central banks want to curb the negative effects of large capital inflows on export competitiveness (as in Romania throughout 2007), but it endorses interventions tailored to protecting private banks’ loan portfolios. This suggests that the IMF’s new crisis economics is unsuitable for countries highly exposed to tensions in international financial markets. A sudden reversal of large capital inflows accentuates the ‘unemployment bias’ (Fontana, 2009) by forcing central banks to accept large declines in output and lower future growth to ensure exchange rate stability. Behind the rhetoric of flexibility, the IMF continues to advise pro-cyclical policies.

Lacking consistency

This is even more puzzling because theoretically, the central bank’s abandonment of strict inflation targeting (in this case by worrying about exchange rates) leaves room for counter-cyclical spending by the government. This fiscal activism would be supported by unconventional monetary policies like quantitative easing – direct purchases of government bonds by the central bank that ensure interest rates on government debt remain at sustainable levels. The IMF’s own research described the positive impact of increased government spending in a recession, further supported by its endorsement of such policies in high-income countries in 2008-9 (Spilimbergo et al., 2009). Yet the institution refused to apply this reasoning in its Eastern European programmes by failing to recognise the importance of coordinating monetary and fiscal policies. Its advice on taxation offers a powerful example: tax increases pushed inflation above targets even though core prices (excluding volatile items such as food or energy) were on a deflationary trend. To retain credibility, the IMF advised tighter monetary policy, effectively curtailing the central banks’ willingness to consider a wider range of policy measures, such as quantitative easing – the approach in high income countries such as the US or UK.

On the contrary, the IMF presented less government spending and “prudent” monetary policy (read high interest rates) as necessary to calm financial markets, a position apparently vindicated by Greece’s sovereign debt problems. However, if any lesson is to be drawn from the Greek experience, it is that the central bank can alleviate debt pressures. When threats of contagion prompted speculation of the eurozone’s collapse, the European Central Bank (ECB) abandoned its long-held reluctance to purchase member states bonds. These ECB purchases immediately lowered the interest rates demanded of European sovereign borrowers (Financial Times, 2010). In turn, Eastern Europe’s experience demonstrates that central banks’ refusal to support sovereign debt forces governments to finance deficits by borrowing on domestic money markets at short-maturities (between one and six months) with high interest rates, necessitating further fiscal contractions.1

Benefitting private finance

The question to ask is who benefits from the IMF’s new design of monetary conditionality? As I have argued elsewhere, the IMF’s advice in Eastern Europe sought to renew policy commitments to private finance, from the insistence on protecting commercial banks’ balance sheets to its organisation in Vienna of the European Bank Coordination Initiative (EBCI) during its negotiations with Romania in March 2009 (Gabor, 2010b). The EBCI brought together what the IMF viewed as key stakeholders in the process of securing Western European bank’s commitments to stay in the region: the parent banks, country authorities and international institutions (ECB, European Commission, European Bank for Reconstruction and Development). The absence of civil society organisations or trade unions (to represent highly indebted households or public sector workers) and the lack of transparency raise questions about the decisions made in Vienna. Indeed, while governments reiterated commitments to austerity, parent banks conditioned their continued presence in Eastern Europe on the availability of “appropriate investment instruments” (Andersen, 2009). From this perspective, the apparent inconsistencies in the IMF’s treatment of central banking can be read differently: ‘prudent’ monetary policy involves maintaining interest rate differentials and liquidity policies that enable commercial banks to obtain high returns in government debt markets or currency markets (Gabor, 2010b).

Hungary’s experience is illustrative. During 2009, Hungary was the IMF’s star pupil in Eastern Europe. Unlike its peers, it remained committed to the original deficit target; its output contraction was lowest, lending weight to the IMF’s insistence on fiscal rollbacks to restore confidence and growth. Elections in May 2010 brought in a new government that promised to reverse the social costs of austerity by asking “the strongest participants of the economy [to] help those who are still in distress” (Financial Times, 2010). First the government imposed, in July 2010, a 0.7 per cent of GDP levy on banks (four times higher than anywhere else) that the IMF decried as “highly distortive” for banking activity – as if banks had all along been governed by the smooth laws of equilibrium rather than by business models that brought the financial crisis to Eastern Europe. The IMF immediately suspended its Hungary programme, predicting a severe market reaction and the possible withdrawal of foreign banks. Such a withdrawal never materialised. While committed to the deficit target initially agreed with the IMF for 2010, in October the new government sought to stimulate private demand by cuts in both income and corporate tax. To maintain revenue it instead imposed temporary ‘crisis’ taxes on businesses in the highly profitable retail, energy and telecoms sectors, and shifted pension funds from private financial institutions to finance the deficit – all measures condemned by the IMF. The Hungarian government further questioned the Hungarian central bank’s policies, demanding lower borrowing costs to boost spending and investment. This very public conflict was aggravated by the central bank’s decision to raise interest rates twice in late 2010, citing concerns with inflation.

What Hungary’s experience demonstrates is that macroeconomic policies, and inevitably the IMF’s conditionality, play a crucial role in deciding how to distribute the burden of adjustment during crisis. The IMF’s new economics of crisis claimed to entail an objective approach, however, the record in Eastern Europe shows that its interventions are mediated by politics and tailored to financial interests. Indeed, the IMF originally embraced inflation targeting because it allowed countries to abandon exchange rate concerns and focus on influencing inflation through aggregate demand control. Yet in response to a crisis triggered by banking sector practices, conditionality continues to be dominated by fiscal concerns, whereas monetary policy advice is quietly geared towards ensuring profitable investment opportunities for private finance. The new economics of crisis legitimises policy advice that imposes ‘antisocial’ measures (wage cuts, public sector layoffs, taxes on consumption) in order to protect financial sector returns.