Sunday, May 24, 2020

Where stands the crisis of 2020? Some insights from a century of leading and lagging indicators

The Irish economy is no stranger to crisis.  Since the 1950s, the most severe of these have occurred in roughly 25-year intervals.  The current crisis has brought about the sharpest downturn, at least in employment terms, that the economy has ever experienced.  As the duration remains uncertain it is cannot be definitively classed as severe so it may be that comparisons to previous crises aren’t wholly appropriate but it may be instructive to look at what preceded and coincided with previous episodes.

Regardless of whether  COVID19 has ended the sequence of 25-year intervals between severe economic crises hopefully we have not returned to the rhythm of the early years of the State when no decade passed without a severe crisis, or more, of some form or another.

The chart below shows a time series for the real growth of the Irish economy since just after Independence. The series is smoothed by taking a three-year, centred average.

Real Growth 1924-2019

The shaded periods are when this three-year average growth rate was negative, and this will also be used in the charts that follow.  Five such negative growth episodes can be identified:

  • 1930s: Great Depression and Economic War
  • 1940s: The Emergency (WWII)
  • 1950s: Balance of Payments Crises
  • 1980s: Procyclical Fiscal Policy
  • 2000s: Bursting of the Credit Bubble

The Gerlach and Stuart (2015) estimates are based on GDP and provide the most plausible annual estimates up to the late 1940s.  A series based on GNI* from Fitzgerald and Kenny (2017) is shown from then but it differs little from the Gerlach and Stuart series right through to the end of the 20th century. 

The Fitzgerald and Kenny series is chosen because it better fits with the constant price GNI* for 2013 on as published by the CSO.  A chart showing unsmoothed versions of both historical series, and illustrating the strong overlap between the two, is here.

The negative growth episodes of the nineteen thirties, forties, fifties, eighties and a decade ago could be joined by a similar outturn for the start of the twenty twenties.  Given that we have gone through periodic economic crises it may be instructive to look for some “then and now” comparisons for some key indicators.

The most severe of Ireland’s economic crises have typically being preceded by a deterioration in the current account of the balance of payments.  From Fitzgerald and Kenny (2017) we have annual estimates of the current account balance as a share of gross national income beginning in 1938 which here is combined with the modified current account of a share of GNI* from 1995 on.

BoP Current Account 1938-2019

As stated, a deterioration in the current account has been a leading indicator for major negative growth episodes, with the deteriorations in the late-1970s and mid-2000s in particular due to economic mismanagement and pro-cyclical fiscal policy.  The presence of a large balance of payments deficit before these crises hit severely restricted the policy options available to respond to the downturn with “restoring order” taking precedence of over supporting incomes and economic activity. 

But in a case of “this time it’s different” we can see that in the past few years there has been a balance of payments surplus that is unmatched since the years of The Emergency (World War II to the rest of the world).  Indeed the 6.3 per cent of GNI* figure for the 2019 current account surplus taken from the Stability Programme Update may be an underestimate.  This strongly suggests we do not have structural imbalances that need to be corrected first before responding to the crisis.

When the slowdown took hold in 2008 one of the first fiscal responses was to announce a package of spending cuts in July 2008.  This time around one of the response was to announce a multi-billion package of increased spending.  A year ago, we said that the economy had savings that could have been spent but suggested it maybe should be a sector other than the government sector that did so. 

We can look at how the sectoral balances contribute to that six per cent of national income surplus we have been running on the current account (with all adjustments to make the sectoral balances consistent with the modified current account applied to the non-financial corporate sector which is where the MNC distortions occur).

Sectoral Balances and the Modified Current Account 2001-2019

A chart showing the domestic sectors for recent years stripping out the impact of foreign-owned companies is here.  Previously, we thought it might be the household sector that could cut loose a bit, but needs must, and it is the government sector that is doing the spending.  Indeed, it could be that the household sector, which has been deleveraging for a decade actually increases its savings during the crisis which is something we will come back to. These savings can be used to fund the government’s spending.

Of course, flows are only part of the story. Stocks matter too.  From Fitzgerald and Kenny (2017) we also have a long-run series of government debt as a share of national income.  The path of the ratio of public debt to national income before each crisis has not been the same.

General Government Debt to GNI 1924-2019

The debt ratio was already rising in advance of the crises of the 1950s and the 1980s.  The ratio was declining in advance of the 2008 crash but that still saw the largest run-up of public debt of any crisis (to date).  In recent years the debt ratio was on a declining path, however the level is still elevated and at around 100 per cent of national income is the highest it has been as the country potentially enters a severe crisis.

For private debt we can get a measure of personal, i.e. household, debt from Stuart (2017) and this shows that increasing household debt was really only a factor in advance of the crash of 2008.

Household Debt to GDI 1949-2019

And in the decade since, the Irish household sector has undertaken a remarkable level of deleveraging – reducing debt was what a large share of the household sector surplus shown with the breakdown of the current account was used for.  Household loan liabilities have fallen from €205 billion at their peak to around €130 billion now.  As the chart shows, the household debt to income ratio peaked at around 215 per cent in 2009; at the end of 2019 it was 115 per cent and is approaching ‘normal’ levels faster than anyone could have imagined.

As stated earlier, it could be that the household sector actually increases its savings in response to the crisis – this may show through an increase in deposits.  This would not be unusual.  The household savings rate is something that tends to react to a crisis rather than something that foretells it. 

Household Savings Rate to GDI 1945-2019

Using Stuart (2017) we have estimates going back to 1945 and it can be seen that the savings ratio does increase at the start of each negative growth event meaning the reduction in household consumption is greater than the reduction in household income.  Remarkably, the Stability Programme Update estimates that the household savings rate will jump from 10 per cent in 2019 to close to 20 per cent this year, which would be the highest for the series by some distance.  In nominal terms this would be an increase of around €10 billion.

The savings rate only reflects the difference between income and consumption; one would need to add the impact of investment spending to get the final non-financial position of the household sector. Such a sectoral breakdown of investment is only available back to 1995 and is what is shown in the chart of sectoral composition of the current account.  The large borrowing position of the households sector in the decade before 2008 is clearly evident, as is the net lending the sector has done in the decade since.

Another response variable or lagging indicator can be the real growth in household disposable income.  Using Stuart (2015) we can see that this followed the same pattern as each of the last three major negative growth crises took hold: it too turned negative.

Household Disposable Income Real Growth 1945-2018

We do not yet know what will happen to aggregate household income in this crisis.  Yes, earnings will fall very significantly for a period but there will be an offsetting effect from government transfers.  As already pointed to, additional social welfare spending of almost €7 billion has been allowed for (thus far).  This is countercyclical policy and is maybe another reason why comparisons to previous crises are not be appropriate.  We still don’t have enough information but if there is a shaded area for the early-2020s added to future versions of these charts it could be the first time the real growth of household disposable income does not also turn negative (on a three-year average basis).

Fiscal policy was a contributory factor to the reduction in household income seen in the 1950s, 1980s and a decade ago.  Some automatic stabilisers would have played a role in mitigating income losses but steps taken to address imbalances, perceived or otherwise, which included tax increases and real expenditures cuts, meant that the impact of the downturns on incomes was exacerbated by fiscal contractions.  Fiscal policy was procyclical.

From Fitzgerald and Kenny we have a long-run series of the balance of the government sector right back to the earliest years of the State.  It can be seen that there was a wide range of levels and trends in the government balance prior to the previous three negative growth events.  There was a deficit that was reducing in the mid-1950s; the late-1970s saw a large government deficit that was increasing and during the mid-2000s budget surpluses were recorded. 

General Government Balance to GNI 1924-2019

But what is as significant is what happened to the government balance not long after economic growth turned negative in the 1980s and 2000s: it improved.  And this was a forced improvement brought about through tax increases and real expenditure cuts which started almost as soon as those crises hit which is the opposite of what is required to support incomes and economic activity in a downturn.

The past few years have seen the headline government balance improve and move to a (small) surplus.  If the government had stuck to its own spending plans maybe the surplus would have been a bit bigger and fiscal policy was probably acyclical, at best.  But there is capacity now for policy to be countercyclical and in large part it is down to what is shown below.

General Government Interest to GNI 1924-2019

At the end of 2019, government debt was equivalent to almost 100 per cent of national income.  In the 96 years since 1924, according to the estimates from Fitzgerald and Kenny there have been only been 13 years when the debt ratio was higher.  But if we look at the interest chart we see that the interest to national income ratio does not reflect this.  In fact, since 1924 there have only only be 35 years when the interest ratio was lower than the 2.2 per cent recorded in 2019. 

As some of this interest is paid to the central bank it is recycled back to the government.  And as long as the ECB keeps the taps opened the massive additional borrowing that will happen will not significantly increase the interest ratio – in the short term at least.

And to conclude, here another lagging indicator that illustrates why this crises will be different to previous ones: the escape valve of emigration is unlikely to be available to alleviate unemployment.

Net Migration per Thousand 1926-2019

This is yet another series where the response to the 25-year crises is clearly evident.  In fact, it has been pointed out that the crisis of 2020 may be the first since that linked to The Great Depression in the early 1930s when net outward migration does not increase.  Emigration did increase in the mid-1930s but by then the international economy was pulling out of The Great Depression. 

Taking the unemployment series from Gerlach, Lydon and Stuart (2016) it is probably safe to predict that the record unemployment from 1935 of 18 per cent will be exceeded in 2020 (at least for part of it).

Unemployment Rate 1923-2019

So where stands the crisis of 2020? The short-term shock is likely to be the most severe the economy has ever experienced.  But relative to what has gone before we have identified a number of key differences. 

The most significant of these is probably the large balance of payments surplus.  The government might not have a rainy-day fund but the household sector has been saving for a decade.  And as the government cuts back on its emergency measures there will be capacity, and hopefully the confidence necessary, for households to increase their spending. 

A second difference arises from the fact that this is a global crisis. If a country using a common currency experiences an asymmetric shock there is a risk of interest rates exacerbating the problem.  Ireland has a high level of government debt but the medium-term risk of adverse interest rate moves is low so we can expect the interest burden of public debt will remain low.

We enter the crisis from a position of structural strength.  And the response so far, at a macro level at least, can be considered to have been appropriate.  Duration remains the key unknown.  But if the phased re-opening of the economy is successful, and microeconomic policy can keep viable businesses alive, it may that this shock, sharp and all as it will be, does not break the 25-year cycle between severe, multi-year negative growth events of the Irish economy.  Here’s hoping.

No comments:

Post a Comment