Tuesday, 12 April 2011

Facing Up to Reality II: The Methodological Flaw in An Bord Snip Nua

This post follows on from a previous contribution.

Michael Taft: In the previous post, we saw how billions of fiscal contraction has led to little deficit reduction. After that post was written the IMF published their latest projections. They estimate the deficit this year to be -10.8 percent this year. Between 2009 and 2011, we have experienced a fiscal contraction of €10 billion - or over 6 percent of GDP. Nominal GDP will fall by €4 billion. The deficit is expected to fall by less than 1 percent. Does the Government get this connection?

In this post we will examine why the notion that cuts equals savings is one of the more pernicious that has come to dominate the debate; why there is a fundamental flaw at the heart of the methodology employed by the Special Group report. With Government ministers threatening more cuts, this is certainly topical.

The Special Group Report used the word ‘saving’ or ‘savings’ 1,096 times. It neatly equated ‘savings’ and ‘spending cuts’ when no such relationship necessarily exists. Fortunately, we have a simulation of the effects of one of the ‘savings’ that the Special Group highlighted: cutting public sector employment.

Flawed Methodology: The ESRI Stress Test

The Special Group recommended that public sector employment be cut by 17,000. According the ESRI model, reducing public sector employment by 17,000 would mean a reduction of €1 billion in public spending – or 0.6 percent of GDP. What would happen?

• GDP would fall by 0.8 percent. So, for every €1 billion cut, the GDP falls by nearly €1.3 billion.
• More worryingly, GNP would fall by 1 percent. That represents an even more deflationary impact.
• Consumer demand would fall 0.5 percent in the first year, rising to 1 percent in the second year. That’s nearly €1 billion cut from consumer spending – putting considerable pressure on domestic businesses.
• Employment would fall by 1.1 percent. In 2009, that would mean a loss of over 20,000 jobs. Unemployment would rise by almost the same amount.

These are all the factors that must be included before we can assess the ‘savings’ to the Exchequer. So what did the ESRI conclude?

• The deficit would fall by 0.2 percent in the first year and 0.1 percent in the second year.

According to the ESRI, the ‘net saving’ to the Exchequer would be 25 percent of the cut, falling to less than 15 percent in the second year. This is because when you factor in the:

• Loss of tax revenue from reduced spending
• Increase in public sector spending arising from unemployment costs
• Decline in GDP/GNP

The gain to the Exchequer diminishes greatly. This simulation – along with measurements for other spending cuts and tax increases – was available to the Special Group report. It was, and remains, the best estimate of the impact of cutting public sector employment. They didn’t utilise it or even refer to it.

The Special Group could have commissioned, through the Department of Finance, other stress-tests regarding social transfers (nearly 40 percent of the ‘savings’ in the Report was due cuts in direct and in-kind social transfers) and Government purchases of private goods and services, which make up approximately a third of spending on public services They didn’t. They have yet to explain why. But that it would have undermined their basic premise – that cuts equals savings – is fairly certain. For its methodology adopted a crude ‘arithmetic’ approach to spending cuts, not an economic one.

When Government ministers proclaim progress on public sector employment reduction, they are, without realising, actually proclaiming very little progress on deficit reduction but significant progress on deflating the economy, driving up unemployment and cutting domestic demand.

But when this realisation hits home – falling growth, continued high deficits – these same Ministers demand more of the same, again not realising that more of the same is likely to produce the same results which produces more demands for cuts until the economy gives out.

It is a vicious circle, legitimated by the false methodology at the heart of the Special Group report. Cuts do not equal savings. But common sense should tell us this – without resort to models and projections. During a jobs crisis, does it make sense to cut employment levels from the largest employer? Why should we be surprised when the result is so dismal?

We are facing into another round of cuts. Employees are now being threatened - with job losses and pay cuts; in the public and private sector. Why? Because past Government ministers either could not or would not subject their policies to economic stress-tests (any comparison with the banking crisis is not co-incidental). They assumed propositions that had little empirical justification. They suffered from ‘escalation of commitment’ – having committed to a particular strategy, they could not extricate themselves when it became clear the strategy was failing.

It is still not too late for this Government to take a step back from the brink. There is still good will towards it. They could adopt a set of transparent and public measurements whereby fiscal options are assessed on the best data available. And on the basis of such informed analysis, adopt the policies that flow from that.

The last thing the Government should do is merely continue failed Fianna Fail policies with only the most cursory of makeovers. If they do, then people will have the right to ask – what was the election for?


Martin O'Dea said...

"During a jobs crisis, does it make sense to cut employment levels from the largest employer?"

Excellent work Michael
Unless I am mistaken though you and others have been making this arguement for about three years on this website; at every turn what you suggested materialised. Seems like being right is only one element in changing opinion. Excellent comment on that Brad DeLong site

"Ideology starts with a particular belief. For example, Conservative ideology starts with the belief that all government activity to manage and regulate the economy is wrong. Then, even when knowledge and understanding and information comes forward which contradicts that belief, the belief does not change, because it is faith based not logic based. Consequently the Conservatives must attack the knowledge and understanding and information as being incorrect. Only by discrediting facts and knowledge can the belief survive.

Of course, the latest manifestation of this is with the Ryan Proposal. The heart of the Ryan Proposal is that contractionary fiscal policy, policy in which government spending is reduced by a greater amount than taxes are reduced will actually stimulate the economy. Serious people really believe this, because the ideology must be correct. As a result we now have the Ryan Proposal, which is to Economics as Creationism is to Science."

Seamus Coffey said...

Hi Michael,

I hope you're not falling into the trap of equating "adjustments" with "cuts" or "savings"! The ESRI model is very useful but it specifically uses a €1 billion cut in expenditure in its analysis. We've had lot of adjustment but very little in the way of aggregate cuts in expenditure, particularly on the current side.

The adjustment have obviously had significant effects on the people who have suffered because of them, but at the aggregate level, spending has not changed significantly. Of course, there is a huge causality issue here, but lets just work through the numbers.

In 2009, gross current voted expenditure was €55.7 billion. In 2011 it is forecast to be €52.8 billion. We have had huge adjustments but they have not led to similar reductions in current expenditure. Even without the benefit of the ESRI model, it is pretty evident that these adjustments will not bring the deficit down if voted current expenditure is down €2.9 billion from its peak.

The current budget deficit was €11.4 billion in 2009. For 2011, the budget day forecast was a current budget deficit of €11.5 billion! We're going nowhere.

Of course the €2.9 billion reduction in voted current expenditure since 2009 has been virtually offset by the €2.2 billion reduction in non-voted current expenditure (mainly interest on the National Debt) in the same period.

Gross voted capital expenditure has been decimated and is down from €9.0 billion in 2008 to a forecasted €4.7 billion this year. The capital account had a deficit of €13.3 billion in 2009. It is forecast to be €6.1 billion in 2011.

Any "improvement" in the deficit has been on the capital side. And again developments on the non-voted side (money to Anglo, promissory notes have had a big impact).

We have had huge adjustments. But on the current side, not only do these adjustments not equal savings, they don't equal reductions in expenditure.

The deficit is simply the difference between income and expenditure. Unless the constituent elements of the calculation change, the outcome cannot change.

Although most of the emphasis is on expenditure, it is important to remember that expenditure did not cause the deficit. There was a €7 billion surplus on the current account in 2007.

We now have a current account deficit of more than €11 billion. This is an €18 billion reversal on the current account but expenditure has risen by €7 billion in that time. Nearly two-thirds of the deficit has been caused by the collapse in tax revenue.

The only way the deficit can be narrowed is through an and/or combination of expenditure cuts and tax revenue rises. These amounts have to be actually changed not simply "adjusted". It is the fiscal deficit, not the banking disaster, that will bring our debt to critical levels.

Paul Hunt said...

@Seamus Coffey,

Many thanks for this very useful input. What we're seeing in the main is a reassignment of government expenditure and a reallocation of taxation to leave us standing still in fiscal deficit terms. Transfers to the unemployed have increased as jobs have been lost and economic output in the private sector lost. Tax increases have sought to replace the disappearance of the unsustainable bubble taxation revenues. Public capex has been hammered to make up the gap.

I fear, though, that your factual observations will fall on deaf ears here. There are none so blind as those who will not see. The underlying microeconomy of the doemstic sheltered sectors is totally dysfunctional, but addressing this would spook a lot of sacred cows - especially those with ideological horns!

Michael Taft said...

Thanks for that, Seamus – as always, a comment with much to chew over. And, no, I never confuse ‘adjustments’ with ‘cuts’. I support adjustments; it’s just that my ‘adjustments’ would move in a different direction (and, therefore, I would argue they would achieve ‘savings’)! But on to your comment.

When we bore into the numbers we begin to glimpse why ‘cuts’ do not necessarily achieve expenditure reductions. For instance, the Government cut social welfare working-age rates by €8 per week to produce a ‘savings’ of €397 million. However, the loss of demand is likely to lead to business closures/cutbacks and, where this leads to job losses or short-timing, we will see an increase in public expenditure through unemployment costs (which goes beyond the dole – it encompasses all unemployment-related spending: medical cards, Back-to-school payment, rent and mortgage interest supplements, etc.). Then there is the second effect when demand falls again, when those impacted by the first cut (those made unemployed or short-timed because of the social welfare cut) inevitably reduce their demand. And when businesses cut their sourcing, there is that downward effect. The downward spiral continues.

None of the above refers to the loss of tax revenue – let’s stick to public spending. Already we see that the €397 million ‘saving’ is not a saving at all because cutting one public programme may lead to increases in another public programme(s).

Keeping this example in mind, let’s move to the larger numbers. You rightly state that current expenditure is down by €2.9 billion (or 5.2 percent). However, there are certain expenditures that are unemployment or demographic driven (old age pensioners, children, etc.). For instance, even with all the cuts in social transfers – and there’s no doubting that payments have been cut hard – social protection expenditure has increased by 22 percent. A big contributor to this has been the social insurance fund deficit, a result of the recession, which the Government now has to subvent. We can cut, but if people insist on getting old, or having children, or getting unemployed, or getting sick, or becoming orphaned, then we shouldn’t be surprised that the net reduction is less than the headline cut.

If we were to take the social protection budget out of the equation, we would find current spending falling by -11.9 percent. That’s a hefty reduction. And this still doesn’t reflect recession-related spending. We can reasonably assume that impoverishment leads to illness, to take one example, which, in turn, reflects higher demand on public services. Trying to cut these services at a time when demand is rising will always be a rolling-a-boulder-up-the-hill exercise.

In the same vein, there is a growing demand on education given our demographics. And, yet, we’re cutting education (over 3 percent over the last two years). In this instance, even if we kept spending static it would equal a cut.
Therefore, when using aggregate numbers it is important to bear in mind the rate of public spending rise without any fiscal intervention – that is, with neutral budgets. This is just as an important comparator as using year-on-year outturns.

Cuts degrade output. Unfortunately, the ESRI model doesn’t tell us where the savings leakage occurs. For instance, cutting public sector employment achieves few savings. Is this because of the associated tax loss or the resulting expenditure rise? When unemployment rises on a 1:1 ratio with employment loss, I suspect it might veer to the latter.

The 2010 and 2011 budgets cut €8 billion from public spending. Yet the deficit doesn’t move. A number of commentators predicted this result. It is calculable and measureable.

There’s one final point – and this is the stat that dare not reveal its numbers. Namely, that there isn’t that much to cut – unless we start closing whole programmes and services. I intended to address this in a subsequent post.

Michael Burke said...


a useful post.

The central fallacy of current (unchanged) policy is that cuts = savings.

Households are nearly always net savers in aggregate. Corporations are usually net borrowers (via the banking system) of those savings for investment. When they stop investing they cause recessions- they too have become net savers. This obliges the government to become a net borrower (as, leaving aside the overseas sector, the household and corporate sector must be lending to another sector).

This inceased government borrowing (dissaving) is an inescapable function of the economy while the other two sectors remain net savers.

Current policy is premised on the notion that if the government tries to reduce it borrowing via cuts the household and corporate sectors will increase their own spending. As many here warned, the opposite occurs, both sectors increase their saving (household consumption and investment are both still falling). The net result is increased government borrowing, through lower tax revenues and increased welfare payments, even as welfare entitlements are slashed.

But the opposite is also true, and proven currently in countries as diverse as the US, China, Germany and France. Increased government spending encourages a rundown of savings in the other 2 sectors, that is increased consumption and investment. With a short time lag this leads to a decrease in government borrowing.

Robert Sweeney said...

@ Michael Taft

It would be useful to calculate the change in social protection spending per recipient/beneficiary. Is that possible? While aggregate social protection spending may not have moved much, surely expenditure per beneficiary has.

Michael Taft said...

Robert - unfortunately, the data on the number of beneficiaries and recipients lag. The latest Department's Statistical Information Report relates to 2009. However, the rates are a good guide. Since 2009, basic rates have been cut by 7.8 percent (excluding pensions). In addition, Child Benefit has been cut by 15.7 percent, Early Childcare Supplement was abolished (€92 per month) as was the Christmas Bonus which represented an additional week, or 1.8 percent; this impacted on old age pensioners. There was also further reductions for recipients under the age of 25. That shows there ere real and identifiable cuts in spending.