Thursday, 19 November 2009

Multipliers then and now

Michael Burke: There is further evidence that fiscal stimulus works. Kevin O'Rourke has (re)posted here a very interesting piece by Barry Eichengreen et al here - on the parallels between the period of the Great Depression and the current 'Great Credit Crisis'. Taking into account the global economy (not just the US) then and now, the authors find close and very disturbing parallels; that global industrial production has fallen as fast now as then, that global stock markets have fallen more now and that the collapse in world trade is even graver now.

The authors also examine the effectiveness of monetary policy then and now, focusing on short-term interest rates. They also examine the effectiveness of fiscal policy changes.

Their key conclusion is also the most interesting and relevant one;

“[Scepticism suggested that] monetary policy is ineffective when the banking sector is in distress. Fiscal policy is ineffective when the need is to reduce the level of indebtedness and when much previous output in the declining sectors is unsustainable; it simply cannot be replaced by replacing demand.

“Our results push back against this scepticism. They suggest that fiscal stimulus made little difference in the 1930s because it was not deployed on the requisite scale, not because it was ineffective” (p.25).

The authors go on to add that their mathematical estimates for the government spending fiscal multipliers were 2.5 in year 1 and 1.2 thereafter.

Now, we are repeatedly told that Ireland is a small very open economy, where the fiscal multipliers would not work in the same way; that government expenditure would leak abroad in the form of increased import demand. Yet many of the leading economies discussed in the Eichengreen paper were large very open economies in that they were centres of vast established political and economic empires, or putative ones. Furthermore, the colonial possessions were the destinations for huge capital outflows and the source of ultra-cheap labour and imports. Therefore, it would be reasonable to assume that the 'leakage' overseas of any fiscal stimulus would be fairly high, certainly no lower than 21st century Ireland.

But we do not have to rely on supposition in that regard. Bang up to date, the government's Pre-Budget Outlook (PBO) tells us what some of the multipliers are for the output that has been lost to date in the Irish recession and their effects on government finances.

The PBO states that, in the past, the official expectation was that a 1% change in nominal growth would lead to a 1.1% change in tax revenues (strangely low by intenational standards, but perhaps reflecting Ireland's very narrow tax base). But in the last 2 years, tax revenues have declined by 32% while the value of GDP has declined 13% (p.20.). That is, a revenue multiplier of approximately 2.5 over 2 years.

Now, no doubt, there will be those that argue that this is previous taxation data based on the housing bubble, as if that was an argument for maintaining the current narrow tax base. But the decline in Ireland's tax revenue is not wholly attributable to housing as there has also been a commensurate decline in other areas of investment, notably machinery and office equipment, which had a peak-to-trough decline of over two-thirds. In any event, it is entirely possible to adjust the tax base to reflect any areas the government chose to target for reflationary measures.

But that's only one part of government finances. The other side is expenditure. Here, the PBO states that there will be a rise of €4.5bn between 2008 and 2010, "predominantly the rising cost of social expenditures due to an increase in unemployment." This represents a rise in spending of approximately 8.5% in response to a decline in GDP of 13%. Even this total significantly understates the effects on spending given the large number of cuts already enacted.

Based on the PBO's current spending total of €56bn in 2009 and tax receipts of €32bn and the effects on government finances to date, a 1% rise in GDP could yield increased taxes of €790mn and lower spending by €350mn to give a total saving of €1.14bn over two years.

The alternative, attempting to cut your way to a balanced budget has already been tried and has already failed, as the PBO inadvertently admits. The increase in unemployment noted above, and the increasing costs associated with it, predictably, "have more than offset the expenditure economies already announced."


Michael Taft said...

It is frustrating that this type of analysis doesn't 'leak' into the wider debate. As to multipliers in Ireland - the Lane/Benetrix numbers show that capital investmnet sarts from 1.2 in the first year, to 1.6 in the second. For Government consumption of market goods/services it hits at over 2 in the first year. As these numbers range over two decades, they would impact higher during a recession. There's little evidence of widespread leakage.

And when we examine the CSO's Input-Output tables, we find that nearly 50% of business imports (which make up most of our overall imports) comes from the four 'modern' sectors - dominated by multi-nationals. These are hardly sectors which any Government would, or even could, try to stimulate.

Unfortunately, we are stuck with a one-hand clapping debate, especially as all the major parties support contraction. Its as if someone put up a sign on the national debate reading 'ideas not welcomed'.

Michael Burke said...

@ Michael Taft

Thanks for that. I have seen the Lane/Benetrix numbers and they are another useful contribution to what is an imperfect science of gauging the multipliers.

I did see that one party has rejected a contractionary fiscal policy, and set out the case for fiscal expansion.

Sinn Fein's 'Road to Recovery' is very welcome and I hope they will be joined by others.

Proposition Joe said...


Your likening of our intricately globalized economy to the imperial scenario of the 1930s doesn't really stands up to scruntiny IMO.

Simply put, they didn't make flat-screen TVs in those colonies.

Instead that trading relationship was much closer to what currently obtains between the first and third worlds: raw materials flow one way, finished goods the other.

Neither do we have the same "ownership" stake in stimiluating the Chinese economy, as the British or French had in the 30s with respect to their overseas possessions.

Michael Burke said...

@ Proposition Joe

Of course all comparisons are lame.

Ireland buys cars, flat screen tvs, etc from all over the globe, (in August it imported almost as much machinery & transport equipment from China as from Britain and N. Ireland). Britain, on the other hand robbed, say, Egypt of its oil, cotton and other produce, and many more countries besides.

Now, it is often argued here and elsewhere that Ireland should cut its wages to improve competitiveness. That is, as other countries recover (most with fiscal stimulus) the aim would be to get their spending to 'leak' to us.

But there were far greater disparities in wage costs between Britain and Egypt in the 1930s than dreamt of here. So how much more of the stimulus would have leaked overseas from the Empire to the colonies? Logically, more than from Ireland currently.

The point of the research from Eichengreen et al is not that the stimulus didn't leak, but that fiscal stimulus wasn't really tried. This is a lesson which seems to have been learnt, however partially everywhere else.

Again, we don't have to suppose. Michael Taft shows, above, that Irish business imports are dominated by MNCs, and therefore largely impervious to 'leakage' from fiscal stimulus.

Proposition Joe said...


But as the other Michael is fond of reminding us, wages make up only a small component of overall competitiveness. So you've got to compare the human capital & productive base of 1930s Britain and Egypt, versus a similar comparison between present-day China and ourselves.

In many ways China has a much more advanced economy than we do. Whereas the semi-literate peasant, the irrigation ditch and the donkey were the mainstays on the Egyptian economy in the thirties ... China in the noughties is up to its oxters in engineering PhDs, next-gen nukes, and electric cars.

An Saoi said...

Michael (B) - I have in a separate post on this website tried to look at the gross tax figure by source rather than by economic model.

One of the problems with the Irish economy was the complete disconnect between taxation and the real economy. Also economic models assume that all parts of the economy moves in tandem The dis-inflation totally undermines all models leaving the normal planning models in disarray. The D of F seem to have little understanding of the basics.

Economic multipliers will work when additional expenditure is targeted effectively, the Germans have proven that.

Michael Burke said...

@ An Saoi

I saw your post and thought it was very useful. I agree,there seems a huge disparity between the economic models and reality, especially when it comes to government finances. The multipliers are inescapable, which is why the arguments against reflation are obliged to take such strange detours in pursuit of Irish Exceptionalism.

@ Proposition Joe

There is no suggestion that the economies of the 1930s are the same as Ireland in 2009, but then neither is Ireland in 1997, which has undergone a series of structural changes.

Therefore the attempt is made, both by cross-country analysis and analysis over time to understand the specific dynamics of the Irish economy now.

In terms of recent history, during the boom final demand in Ireland grew by 38.4% while import demand grew by 41.6% (2003-07, CSO). There is no evidence from the data that Ireland has a exceptionally high propensity to import. In fact the difference between the two growth rates is marginal.

Niall Douglas said...

"I agree,there seems a huge disparity between the economic models and reality, especially when it comes to government finances. The multipliers are inescapable ..."

I feel a need to make a correction in the technical use of multipliers - and this is to everyone who has referred to them, not just Michael.

Whenever you see "multipliers" in the context of economics it means that one has taken some historical data, made some assumptions to allow the association of some change with some effect across multiple periods, and then one takes the average ratio between changes in the "cause" and in the "effect". This is standard statistics though using much stronger ceteris paribus assumptions than anyone in Biology or Physics would ever use.

Statistics, when used like this, only has predictive power when the structure within the assumed boundaries does not change. Generally, with multipliers, this will tend to hold true for unstable periods of time for large economies like the US and the UK because the boundaries are wide enough and therefore the assumed system is big enough.

The Irish Economy, on the other hand, has relatively little to do with the island of Ireland itself, so one needs to be extremely careful when assuming that the relations between Ireland and the wider trading ecosystem in which it is embedded will be similar in the near future as the near past. In other words, one cannot treat the Irish economy independently of what is happening in Britain, the EU or the US, each of whose economic structures have initiated profound changes recently as well.

This is not to say that using multipliers to point out inconsistencies in rationales for policy changes isn't a good idea, rather I am saying that all econometric modelling inherently requires very strong assumptions which tend to not hold true for extended periods of time.


Michael Burke said...


"all econometric modelling inherently requires very strong assumptions which tend to not hold true for extended periods of time."

I agree. You could also add, or even in the short-run, where there have been sudden changes to the composition of economic activity. Or that some of the assumptions are not always evidence-based.

But we know that some type of multiplier exists with regard to the current situation in Ireland, as with the 13% decline in GDP and 32% decline in taxation revenues. And from the fact that the government keeps cutting and the deficit keeps rising.

The argument here is that we need to get those multipliers (of whatever dimension) working positively for the Irish economy.