The European debate on fiscal austerity has gone astray – focusing exclusively on the size of deficit reductions. What policy makers should really be focusing on is the budget tightening’s composition (tax versus spending) and on the accompanying policies. Indeed, the title of this Vox debate – “Has austerity gone too far? [1]” – reflects this inappropriate emphasis on size.

In our view, the essential question is not 'how far' governments go but of 'how' they go far enough.

Evidence on new taxes versus new spending cuts

Economists have engaged in some lively debates about how to measure and evaluate the effects of large fiscal adjustments episodes in OECD countries (Europe in particular). But a careful and fair reading of the evidence makes clear a few relatively uncontroversial points, despite the differences in approaches. The accumulated evidence from over 40 years of fiscal adjustments across the OECD speaks loud and clear:

• First, adjustments achieved through spending cuts are less recessionary than those achieved through tax increases.
• Second, spending-based consolidations accompanied by the right polices tend to be less recessionary or even have a positive impact on growth.

These accompanying policies include easy money policy, liberalisation of goods and labour markets, and other structural reforms.

There remains a lot of work to be done on identifying the appropriate accompanying policies and understanding the channels through which they help spending-based stabilisations, but the fact is there, as shown for instance in a recent paper by Roberto Perotti (2011).

• Third, only spending-based adjustments have eventually led to a permanent consolidation of the budget, as measured by the stabilisation – if not the reduction – of debt-to-GDP ratios.

IMF research on the austerity composition issue

Two recent IMF publications (IMF, 2010, Chapter 3, and Devries et al 2011) agree that spending-based adjustments are indeed those that work – but not because of their composition, rather because almost ‘by chance’ spending-based adjustments are accompanied by reductions in long-term interest rates, or a stabilisation of the exchange rate, the stock market, or all of the above.

This line of argument is flawed on purely logical grounds. Financial prices – interest rates, the exchange rate, the stock market – are not exogenous. They respond to fiscal policy announcements. For instance, if investors perceive, correctly, that only spending-based adjustments will lead to a permanent consolidation of the budget, this will increase ‘confidence’ and result in lower interest rates and higher stock prices.

A more convincing piece of evidence comes from a comparison of the effects of different ‘types’ of fiscal adjustment on confidence and on output. Tax-based stabilisations not only eventually fail, in the sense that they are unable to stop the growth of the debt-to-GDP ratio. When these fiscal packages are announced entrepreneurs’ confidence falls sharply, and this is reflected in a fall in output. On the other hand, spending-based stabilisations (especially if accompanied by appropriate contemporaneous polices) do not negatively affect economic confidence contemporaneously. Moreover they are often accompanied by an increase in output within a year.

It stands to reasons that European countries where tax revenues are close to 50% of GDP do not have the room to increase revenues even more.

A paper by Harald Uhlig and Mathias Trabandt (2012) nicely shows how close many European countries are to the top of realistically measured Laffer curves. Thus any additional tax hikes would lead to relative low increases in tax revenues and could be very recessionary, through the usual supply- and demand-side channels.

Given all of the above we should stop focusing fiscal policy discussions on the size of austerity programmes. A relatively small tax-based adjustment could be more recessionary than a larger