If media reports are to be believed, the government looks intent on pursuing a large, tax-led fiscal adjustment on April 7. While determination to take resolute action in the face of the recent exchequer returns is understandable, I am concerned that both the size and composition of the proposed adjustment will deepen the risks facing the economy.
(1) Size. My first concern is that the planned adjustment is too large. As pointed out by Jim O’Leary in Friday’s Irish Times, the adjustment is equal to a massive 3.5 percent of GDP in an already reeling economy.
The government faces an unenviable risk-management tradeoff. On one side, it must reduce the risk of unstable public finances. An out-of-control deficit – with the contingent liability of the bank guarantee looming in the background – increases the risks of higher interest rates, a sudden funding stop, or an even more dramatic forced adjustment later on. On the other side, a fiscal adjustment of this size runs the risk of intensifying the vicious cycle of contracting economic activity and falling willingness/ability to spend.
If we assume the 9.5 percent deficit target was appropriate before the news of weaker than anticipated exchequer returns, then continuing to aim at this target means the risk-balancing response is focused only on the public-finances element.
This strikes me as extreme. The rationale appears to be that the government has tied its credibility to keeping the deficit in single digits. I think this stance would be correct if there were solid signs that domestic demand is stabilising. As it is, a too-firm commitment means the government is locked into a potentially increasingly contractionary fiscal stance as the economy spirals downwards.
(2) Composition. Again based on media reports, it appears that tax increases and capital investment cancellations/deferrals will bear the brunt of the adjustment. Alesina and Perotti (1997)* refer to this as a Type-2 adjustment. In contrast, a Type-1 adjustment focuses on cuts to the government wage bill and transfer payments. Although credibly establishing causality with available data is challenging, the international literature has quite strongly concluded that Type-2 adjustments are both less likely to be sustained and more likely to be contractionary.
The evidence points to both demand-side and supply-side explanations for the relative superiority of Type-1 adjustments. On the demand side, Type-1 adjustments have a more positive effect on expected future incomes and on interest rates. On the supply side, these adjustments tend to be more effective in reducing unit labour costs and thus in improving competitiveness. Added to these effects, I believe there is reason to worry that a shift towards higher marginal tax rates will undermine future growth potential and lead to large deadweight losses. (Even the relatively innocuous-sounding removal of the employee PRSI income cap will raise the marginal tax rate of those earning more than 52,000 euro by four percentage points.) And, assuming tax rate increases reach deep down into the income distribution, there is a danger of higher equilibrium unemployment as a result of rising income replacement rates.
Is there a better way? With a large tax-led adjustment as the default, is it possible for a union-supported “grand bargain” that is better grounded in what we know about successful fiscal adjustments?
I believe this would be a more modest Type-1 adjustment combined with a well-specified medium-term fiscal consolidation plan. Interestingly, this is likely to have a good deal in common with Ireland’s successful 1987-89 adjustment.
I have no illusions about how politically difficult this would be in the poisoned aftermath of the pension levy. However, the impressive NESC report gives some reason for hope, especially its emphasis on the power of shared understanding:
“[W]hile a fair sharing of burdens is critical to overall success, the politics that can ensure a sense of fairness (such as taxation and holding financiers to account) will not, in general, be very effective in addressing the economic crisis of firm closures and unemployment. In this sense, the paper seeks to draw on the Council’s particular contribution to Irish policy – the willingness and ability of diverse actors to engage in honest, joint deliberation and analysis of evidence and explanations . . . It is the shared analysis that warrants an integrated response, not the joint engagement that warrants an agreed response.” (p. 5)
I urge the key actors to examine the literature on successful fiscal consolidations in forming this shared understanding. I believe it would considerably lessen support for a Type-2 adjustment.
Possible elements of a Type-1 adjustment adapted to Irish circumstances:
A 10 percent nominal wage cut and (at least) a freeze in transfer payments in real terms
- Cancellation of capital expenditure projects where they fail a cost-benefit analysis on current information. Deferral of projects only when there is substantial doubt about net benefits given the uncertainty about Ireland’s future trend rate of growth. Projects with a high certainty of positive long-run net benefits should be funded now.
- The announcement (with as much detail as possible) of phased plans to broaden the tax base (including a residential property tax and a carbon tax)
Announcement of plans to raise income tax revenues while limiting increases in marginal tax rates to the greatest extent possible (e.g., phasing out tax credits and child benefit payments at higher incomes)
Removal of tax shelters of unproven effectiveness that contribute to a sense of unfairness
If necessary, introduction of a Obama-style higher third rate of income tax that raises marginal tax rates only in the thin part of the income distribution
Announcement of plans for comprehensive pension reform to deal with what is now a major source of social insecurity
*Alesina, Alberto, and Roberto Perotti (1997), “Fiscal Adjustments in OECD Countries: Composition and Macroeconomic Effects,” IMF Staff Papers, Vol 44, No. 2 (June), pp. 210-248.
11 replies on “A Grand Bargain?”
Maybe its because im a time series guy but reading the paper I would be very very cautious about making strong policy inferences at all, and for now, from samples as small as 16 obs , and they over 40 years ago.
While caution is the right approach to any econometric work, I think your characterisation here is quite misleading.
First, the number of observations (i.e. significant discretionary fiscal contractions) in the dataset for the paper is 62 not 16 (the 62 are identified from 378 usable observations in total). 16 is the number of “successful” fiscal consolidations based on a quite stringent condition: “A period of tight fiscal policy is successful if (1) in the three yeas after the tight period the ratio of the cyclically adjusted primary deficit to GDP is on average at least 2 percent of GDP below the last year of the tight period or (2) three years after the last year of the tight period the ratio of debt to GDP is 5 percent below the level of the last year of the tight period.” (p. 221)
Second, you say dismissively that the observations are over 40 years old. More accurately, the data iare based on fiscal policy in OECD countries for the years 1960 to 1994, a period rich in fiscal consolidations (the paper was written in 1997).
Third, this is just one paper in a significant literature. I only referenced this particular paper because of the useful distinction between Type 1 and Type 2 adjustments. Other papers use larger datasets, different definitions, and more demanding econometrics. (See, for example, Ardagna, Silvia (2004), “Fiscal Stabilizations: When do they Work and Why,” European Economic Review, 48, pp. 1047-1074.)
Fourth, there is also a complementary times series literature utilising vector autoregressions. Of most immediate relevance is the recent work by Agustin Benetrix and Philip Lane on the Irish experience (as reported by Philip in his crisis conference paper). They find large positive multipliers for government investment and non-wage government consumption. In contrast, they find a large negative multiplier for wage government consumption. This points in the same direction in terms of the relative effectiveness of Type 1 adjustments.
Thank you for this useful reference.
It seems to me the relative efficacy of Type-I versus Type-II fiscal adjustments will depend on a comparison of the marginal productivity of labour in the public sector to the rest of the economy. Where MPLpub > MPLecon, (and this is usually the case, due to incentives etc.) employ Type I. Assuming f'(L)>0, f”(L)<0, we should eventually reach a point where MPLpub=MPLecon, at which point any further Type-I strategy will cause the economy more overall damage than would an across-the-board hike in taxes. So then proceed with Type II, no?
The planned adjustment is very large but a target deficit of 9.5% of GDP was fixed as a balance between two factors:
(a) the Government did not want to frontload the 16Bn. adjustments needed to get back within the 3% SGP reference figure (politicians always postpone the evil day)
(b) it would be impossible to ask our Eurozone partners to accept a double-digit deficit even for one year.
By this logic, it remains essential to do everything possible to meet the 9.5% target, otherwise we will lose the confidence of our Eurozone partners and the markets in general.
Whether this target is feasible, even with a supplementary budget, will depend largely on the state of the economy and the trends in tax revenue i.e. how a -6% GDP growth rate this year, compared with -4.5% projected in January, will impact on tax revenues. Colm O’Loghlin has a table in today’s Business Post which look like the fruits of some revenue modelling, possibly from his former colleagues in Merrion St.
This is an interesting way to frame it. As I noted in the post, there are potential supply- and demand-side explanations for the superiority of Type 1 adjustments.
Your analysis captures the essence of the supply-side explanation.
We can think of resources being fully employed as a rough approximation for the supply-side effect. If the marginal product is higher in the private sector (I think you meant your inequality to be reversed), then moving resources from the public sector to the private sector would raise total output. In some of the literature, the complementary focus is on how a larger public sector affects the competitiveness of the private sector. Possible channels include higher wages costs (due to direct competition with the public sector) and higher taxes (to pay the public sector wage bill).
But there is also a demand-side argument for the superiority of the Type 1 adjustment. Here we definitely move away from a world of fully employed resources. In such a Keynesian world, we would typically expect both types of adjustment to be contractionary. The additional element to the standard Keynesian story is the effect of the adjustment on expectations of future after-tax incomes and also on interest rates. To the extent that these effects are stronger expenditure-led adjustment — possibly because it is believed to be more likely to be sustained — then the contractionary effect is attenuated.
I can’t disagree that a 9.5% deficit is shockingly large. But neither can I see how it is in the interests of our eurozone partners for us to drive the economy into the ground with an additional 3.5% of GDP in discretionary adjustment. As I think about allowing a double-digit deficit, I worry more about the potential impact on creditworthiness than about the SGP (but I may well be lacking imagination here).
On the creditworthiness constraint, I think achieving the (politically very difficult) expenditure-led adjustment would be a boon to credibility. This, combined with a credible medium-term fiscal strategy that doesn’t do too much damage to supply-side potential, could give the government room to run a modestly higher deficit. No easy options.
Question: “A nominal wage cut of 10% and (at least) a freeze in transfer payments in real terms.” Given inflation is negative and will continue negative for at least this year, are you recommending a nominal decrease in transfer payments? If so, that is brave, but may be politically difficult. If instead you mean a freeze in real terms, but without a nominal decline, that is not a freeze in real terms. Could you clarifY?
Greg, sorry for not being clearer. With negative inflation, I am proposing a nominal decrease in transfer payments. I certainly recognise this would be politically difficult, but I don’t think it is so brave. It effectively allows for the same real transfer payments that were intended on budget day. I believe this is consistent with the critical goal of protecting the vulnerable. There is the question of whether it goes far enough. With private sector real wages declining and taxes rising, this policy stance would still be associated with rising replacement rates, which could have a long-lasting impact on the unemployment rate.
One complication is deciding which price index to use: the CPI or the harmonised index. The rate of decrease is significantly greater with the former, mainly because it includes mortgage payments. There is an argument for using the harmonised index since many on social welfare will not have benefited from a fall in mortgage rates.
I share John’s reservations about the scale of the adjustment that is being planned for April 7th. It might be comforting to believe that the confidence-boosting effect of the kind of heroic action that is being planned will dominate its direct impact on demand, but I don’t share that belief.
I’m hoping (perhaps forlornly) that the reason the government is evincing such determination to deliver a 9.5% of GDP deficit is that is has been given grounds to believe that if it does so, it will receive some sort of benign assistance from Europe, perhaps in the form of an agreement to joint issuance of eurozone bonds.
I fervently hope they’re not planning April 7th because they’re running scared of the signals currently being transmitted by panic-driven international bond markets and their derivatives. I’m especially thinking of the market in CDSs which is widely regarded as dysfunctional and which some finance professionals think should be abolished. On the topic of CDSs, a recent commentary from Davy’s bond analyst, Donal O’Mahony is worth a read.
I also share John’s concerns about the likely composition of the April 7th package. It looks like big tax hikes are on the cards, perhaps with a view to bringing ICTU back on side. There’s nothing wrong with having ICTU on side per se, and we all recognise the need for some tax increase, but there’s a balance to be struck here and it’s not as if the October budget (remember it?) hasn’t already significantly increased taxes. Perhaps, we are partly to blame for the way things are shaping up. In our attempts at even-handedness, some of us (myself included) have perhaps conceded too much ground in relation to taxation.
I’m not clued in enough to know macro economic formula: I do know that adding to a forecast deficit of 6% plus possible deflation of 3 %, by taking out 3.5% in a emergency budget = a possible 12.5% reduction in economic activity. And that scares me.
There is an urgent need to deal with waste in public expenditure- €26m in Dept of Ag travel expenditures etc- but there is also a desperate need to keep (only) important capital expenditure going including, for example, the Ballymun renovation.
Jobs, jobs, jobs.
Can anyone tell me what it costs to keep a reasonable earner on job seekers benefit? At roughly €208 a week plus other state supports it has to be close to €13,000 pa. Let’s keep people working on long term infrastructure. Let the slack at Bord Bia, Dept of Education, Agriculture, LEADER etc move on.
In my view the best approach to defining the appropriate size of the budgetary action should be based on restoring (i) a less cyclically sensitive tax base; (ii) zero structural deficit within 2-3 years. The important difference between this perspective and one based on targeting a particular deficit for 2009 is that it removes the pro-cyclicality of the latter. As for the composition of the changes, it would be hard to disagree with John’s prescriptions. But maybe just as easy to look back to the composition of tax and spending in Ireland before we went of the rails as a useful reference point.
I think Patrick’s suggestion to target the structural budget deficit is a very good idea. Of course, the challenge would be to credibly measure that deficit, particularly given the fog surrounding the potential growth path. It is also critical that the measurement is not politicised. Sounds like a job for the ESRI.