Speevr logo

Global Letter: The economic cost of recessions

Table of Contents

Recessions not only reduce output: they also reduce productive potential

It is a good moment to reflect on the teachings of those with long experience of recessions. One master was J. C. R. Dow, 1 in his definitive Major Recessions: Britain and the World, 1920-1995. 2

Dow was fascinated by what may be termed ‘the anatomy of recessions’. One area of interest among many in his tome was whether – or to what extent – downturns reduced an economy’s potential output. There are various reasons why that might be expected, but principal amongst them was the possibility of investment being lower – and perhaps permanently so – than it would have been had there been no recession. 3

The way that Dow investigated this matter was ingenious. Basically, what he did was to examine the evolution of ‘constant-employment’ GDP – a variable that he constructed – relative to an extrapolation of trend GDP growth during the previous cycle.

Adjusting the figures for actual GDP to a ‘constant employment’ basis was crucial. Unless this is done it is not possible to distinguish a change in GDP that derives from a change in capacity (which is the object of the investigation) from that which results from change in the intensity with which that capacity is utilised.

Dow found strong inferential evidence not only that UK recessions did indeed lead to “downward displacement of the path of capacity growth”, but also that such had been the case following each of the five recessions individually since the 1920s. 4 On no occasion did UK GDP ever reach its extrapolated prior trend: rather, while it asymptoted towards its prior path, it levelled out some way below it. 5 This can be seen in the figures below that, based on our own calculations using the Dow method, replicate his original diagrams.

And so to magnitudes

The Dow calculations suggest that the 1920-21 recession displaced UK productive potential capacity downwards by 4.5%; the 1973-75 recession by 2.2%; the 1979-82 recession by 5.4%; and the 1989-93 recession by a substantial 8.4%. 7

We showed a few weeks ago 8 that it typically takes 3 to 5 years for GDP to re-attain its prerecession level. Add to that the permanent (future) cost of foregone productive potential and it is evident just how expensive recessions are.

In a subsequent Global Letter we shall extend the Dow analysis to include the UK recessions that came after the publication of his book and his death, and also compare the results with calculations using other methods. And we shall also extend the analysis to recessions in some other major economies.

Global Letter: The economic cost of recessions 1

Technical note


The aim of the exercise was to replicate, acceptably closely, the calculations performed by Dow (1998), and hence his two-part Figure 11.1.


GDP t * = GDP t (1 + (u t – u*)), where

GDP is constant-price GDP

GDP* is constant-employment GDP

u is the rate of unemployment

t denotes the year

u* is an arbitrary level of unemployment, which Dow took as 9.8% for the interwar period, and 1.8% for the 1970-93 period.

Data sources

GDP. Various available sources of historical data, including Maddison, the University of Groningen, 9 MeasuringWorth, 10 and the Bank of England. 11 Inspection revealed that the series for real UK GDP from these various sources exhibit near-identical trend changes over a run of years, but on occasions exhibit some, generally small, year-on-year differences.

  • For the pre-war period (1920 to 1938) we used the Bank of England Real GDP, Constant Prices, Chained, Factor Cost, via Macrobond, as there were gaps in the Maddison data.
  • For the entire period 1970-93 we opted for the Office of National Statistics series Real GDP, Constant Prices, Chained, Market Prices, via Macrobond.

Unemployment rate

As for GDP, for the pre-war period (1920 to 1938) we opted for the Bank of England series, extracted using Macrobond.

For the post-war period (1970-1993), we used data from the Office of National Statistics from 1972 onwards, and Bank of England data for 1970 and 1971, which were not available from the ONS.

Plotting the data and trendlines

GDP and GDP* are plotted as log-levels, so that a straight line depicts a constant rate of growth

Trendlines. Dow’s trendlines ran from the end of a given recession until the start of the next one, thus:

  • Fitted 1921 to 1929, and extrapolated to 1938.
  • Fitted 1932 to 1938.
  • Fitted 1970 to 1973, and extrapolated to 1993.
  • Fitted 1975 to 1979, and extrapolated to 1982.
  • Fitted 1982 to 1988, and extrapolated to 1993.

Dow reports that he “fitted” these trend lines. For our part we simply drew straight (log linear) lines that went through the beginning and end points, and for all practical purposes Dow’s trend lines and ours are the same.

  1. Christopher Dow was variously a Senior Advisor to the U.K. Treasury, Deputy Director of the National Institute of Economic and Social Research, Assistant Secretary General of the OECD, and Executive Director of the Bank of England.
  2. Dow, C., 1998. Major Recessions: Britain and the World, 1920-1995. Oxford University Press.
  3. In that respect, Dow’s analysis was more fundamental than is implied by the currently-trending, rather nebulous, term ‘scarring’. It is hard to imbue that term with a precise economic meaning: perhaps, at best, it is an allusion to the tendency for workers’ skills to atrophy when they are out of work for a protracted period. That is without doubt a real issue, but it is probably quantitatively less important than the foregone-investment effect.
  4. And this phenomenon clearly remains alive today. To take just one example, the International Energy Agency (IEA) has recently reported, on the basis of tracking corporates’ investment intentions, that investment globally in energy production this year (2020) stands to be some 22% below what it would have been but for the coronavirus recession. See IEA, 2020. World Energy Investment 2020: Key findings. May. Available at [Accessed 28 May 2020]
  5. In Dow’s words: “… a major recession of demand …. result[s] in a downward displacement of the path of capacity growth … i.e. the path of full employment (when, if ever, regained) is on a lower level – though on a path that may be parallel to a projection of the previous capacity growth path i.e. some economic capacity is permanently lost.” Dow (1998), op. cit., p. 174
  6. See Dow (1998), op. cit., p. 386. A technical note detailing how we performed the calculations is provided in the annex.
  7. Taken together, these losses represented, according to Dow’s calculations and noted in his Figure 11.1, a 23% reduction vis-à-vis the (extrapolated) 1955 to 1972 trend. However, it is debatable how appropriate it is to apply such a calculation to a very long period: it is not necessarily plausible, for example, to take, as the constant-employment GDP trend, for the whole period, the (relatively high) rate of growth that obtained just in the period of immediate post-WWII reconstruction. Nevertheless, Dow’s basic point stands: irrespective of what trend or trends are taken, the losses of potential output from successive recessions cumulate to an extremely large figure overall.
  8. See Llewellyn J., Jones, R., and Park, S., 2020. Recessions last years, not quarters. Llewellyn Consulting, 7 May.
  9. The original Maddison historical database has been preserved by the University of Groningen Growth and Develop centre, which also continually extends and develops these data further. For more, see University of Groningen Growth and Development Centre, available at [Accessed 31 May 2020]
  10. Thomas, R., and Williamson, S., 2020. What Was the U.K. GDP Then? MeasuringWorth 2020. Available at:m http://www.measuringworth.com/ukgdp/
  11. The Bank of England’s A millennium of macroeconomic data is the result of a large exercise to bring together a range of data sets and integrate them to the greatest extent possible. For more see Bank of England (2020). Research datasets: A millennium of macroeconomic data. Available at [Accessed 31 may 2020]
Subscribe to receive more by Llewellyn Consulting

Most recent by


Let's get started...

Speevr Logo

Share this page

Global Letter: The economic cost of recessions

Dow was fascinated by what may be termed ‘the anatomy of recessions’. One area of interest among many in his tome was whether –