Abstract

Current economic turbulence has revived interest in interwar macroeconomic instability and policy. This paper provides a guide to the current state of knowledge on British macroeconomic policy between 1929 and the eve of the Second World War for today’s economists and policy-makers. Mindful that some will seek ‘lessons’ from this earlier age, it makes clear the very particular economic and policy context of the time, including the marked difference between British and American economic performance and policies and the challenge posed by the massive rise in debt brought about by the First World War. Knowledge transfer is possible here for policy learning, but what transpires is that lessons are typically stronger as negatives than as positives: namely what should not be done in general, as against what might be done, this largely situation specific. The focus is the role and effectiveness of the monetary and fiscal policies pursued together with an assessment of the principal policy not adopted, namely the Keynesian solution of significant loan-financed public works to remedy mass unemployment. The analysis is comparative with respect to US macroeconomic economic performance and policy; it explores the different macroeconomic shocks of the ‘contraction’ phase (the term ‘Great Depression’ is not really appropriate to the British experience) in the two countries; comparative policy regimes; and then, in turn, analyses the policy mix and policy effectiveness of monetary and fiscal policy, including for Britain the counterfactual Keynesian solution. A business-cycle periodization is adopted, with the key role for the abandonment of the gold standard in terms of regaining freedom of choice in macroeconomic policy, itself experiencing significant developments after 1931. The traditional role of the cheap-money-induced housing boom in Britain’s post-1932 economic recovery is reaffirmed; the impact of the fiscal policies actually pursued are then discussed, as are assessments of the likely effectiveness of the Keynesian solution had the 1929 Liberal Party’s ‘We can conquer unemployment’ spending package been implemented.

Introduction

Explaining why the Great Recession beginning in 2008 has not, at least to date, become a 1930s-style Great Depression in Britain, or indeed elsewhere, is rightly a current preoccupation of economists and policy-makers. In Britain, as in the US, the government budget deficit deteriorated very substantially between 2007 and 2009 (by some 8 per cent of GDP), with the greater part of this due to fiscal deterioration rather than the cost of bank support (Holland et al., 2010, Figure 1). In Britain, public-sector net borrowing in the 2009 calendar year reached £159.2 billion, some 11.4 per cent of GDP (ONS, 2010), and much was made of this figure as a peacetime record, not least during the prolonged general election campaign.1 Amid the welter of comment on the present situation, the historical record has become variously a rhetorical weapon to be deployed or a knowledge base to be scoured for appropriate lessons. ‘History’ is here invoked typically in two distinctive senses: the first as a risk of repeating, or the prospect of avoiding, through policies unwisely or wisely chosen, a macroeconomic catastrophe of 1930s duration and depth; and the other as either the worst fiscal imprudence of any post-war British government or the inevitable and justifiable consequence of a responsible government allowing the automatic stabilizers to operate to mitigate depressive impulses. In the process, of course, there has also been a major revival of interest in Keynes and Keynesian stabilization policies.

Figure 1

Real GDP (1929=100), US and UK, 1929–40

Sources: UK: Sefton and Weale (1995, Table A3); US: Carter et al. (2006, III, series Ca9).

Figure 1

Real GDP (1929=100), US and UK, 1929–40

Sources: UK: Sefton and Weale (1995, Table A3); US: Carter et al. (2006, III, series Ca9).

In this paper we examine British monetary and fiscal policy from the 1929 downturn to the eve of the Second World War from the perspective of today’s economist and policy-maker needing to understand the specificity as well as possible knowledge-transfer potential of this earlier key policy episode. The Federal Reserve chairman, an economist whose research was the Great Depression before he became a policy-maker,2 has recently identified four policy lessons from history: first, ‘economic prosperity depends on financial stability’; second, ‘policymakers must respond forcefully, creatively and decisively to severe financial crises’; third, ‘international crises require an international response’; and, fourth, ‘history is never a perfect guide’ (Bernanke, 2010). He does not accord weights to these lessons, but as we turn now to consider Britain in the 1930s we here signal that on the basis of what we now think we know about British policy, the major lessons, and here we assume we can surmount the epistemological problem of knowledge transfer, are stronger as negatives than as positives, namely what should not be done in general as against what might be done, this largely situation specific.

In assessing British monetary and fiscal policy in the 1930s we have a number of objectives. First, we consider the role and effectiveness of the policies pursued as well as the principal policy not adopted, namely the Keynesian solution of significant loan-financed public works to remedy mass unemployment. Second, wherever possible the analysis is comparative with respect to the US and seeks to highlight what is specific to the case of the 1930s and what might be generalizable. This involves consideration of one of the biggest questions for both the 1930s and the current situation: that of policy-makers’ tolerance or not for debt and the judgements made accordingly on whether the automatic stabilizers can be allowed to operate or whether they should be overridden through discretionary adjustments to taxation and expenditure. In Britain in the 1930s, the authorities not only eschewed the Keynesian solution but, from the outset of the downturn, sought to override the automatic stabilizers. By contrast, for today’s policy-makers the key question has been about the desirable path for fiscal consolidation: neither too fast nor too deep, lest recovery be threatened; nor to slow and too shallow, lest the markets lose their tolerance for sovereign debt. Of course, since the Great Recession has not become a Great Depression, and, in particular, unemployment to date has not risen as much as feared initially, the question arises of whether it was policy-makers’ tolerance of the high levels of debt creation compelled by the massive budget deficits that prevented a slump. There is thus a clear counterfactual: an earlier fiscal consolidation, one which overrode the automatic stabilizers, would have produced a sharper and more prolonged downturn. Indeed, this was the rationale of the outgoing Labour government’s position on fiscal consolidation, one maintained in relation to their opposition to the new Conservative-led coalition government’s emergency June 2010 budget which accelerated the path of consolidation and altered the balance as between taxation and retrenchment. As today’s policy-makers grapple with this issue, the example of Britain in the 1930s is of major interest: fiscal consolidation, involving both increased tax rates and expenditure retrenchment, was achieved (albeit quickly interrupted by the exigencies of rearmament), but arguably at the cost of enduring high unemployment.

The structure of this paper is as follows. Section II situates the interwar British economy in terms of some stylized facts which condition policy priorities and responses. Section III sets the scene in a different way, by examining the policy regime which was importantly very different from today. Policy is then the focus of section IV, which considers in turn monetary then fiscal policy, including the Keynesian solution not adopted. Section V provides an overall assessment and concludes with a summary of the lessons of Britain in the 1930s for today’s economists and policy-makers.

Britain in the 1930s: some stylized facts

A necessary starting point is with a UK–US comparison of economic trends, from which it is clear that the British economy did not experience a ‘Great Depression’ in any meaningful sense between 1929–32. Similarly, its ‘recovery’ was of a different order of magnitude with, relative to the US, greater scope to argue for the positive effects of government macroeconomic policies, even if they did fall far short of what Keynes and other economic progressives considered desirable and deliverable.

Using annual data, we chart in Figure 1 the relative paths of real GDP on a 1929 base: these show a peak-to-trough fall in the UK (1929–31) of 5.4 per cent and for the US (1929–33) 26.6 per cent, with the respective 1937 values being 16.4 per cent and 5.3 per cent above their 1929 levels. No wonder, then, that from an international perspective Schumpeter, for one, questioned in 1939 whether Britain experienced a great depression in any meaningful sense (cited in Richardson, 1967, p. 17). Hatton and Thomas (2010) discuss comparative labour-market trends, and here we just note that average unemployment was much higher in Britain than in the US in the 1920s, giving very different starting points for 1929, but that, relatively, the UK record was much better after 1932.

In terms of the downturn, we do now have higher frequency national account data for the UK and this gives a more pronounced peak-to-trough loss of between 7.6 per cent (quarterly data) and 8 per cent (monthly data), and a trough-to-peak rise of 26.5 per cent on both series,3 but we lack these data for the US. Even so, the relative mildness of UK income losses remains, as does the pronounced recovery after 1932. However, the UK output–employment elasticity was relatively high (Middleton, 1981, p. 277), such that the contraction phase of the real economy was manifest in a significant rise in unemployment and fall in employment: labour-force unemployment rose from an annual average of 1.5m (7.3 per cent) in 1929 to 3.4m (15.6 per cent) in 1932 and was still at 1.8m (7.8 per cent) by 1937 (Feinstein, 1972, Table 57). The figures used by contemporaries, a by-production of the national insurance system which covered only 60 per cent or so of the labour force, gave much higher unemployment rates, peaking at 23 per cent in January 1933,4 a peak broadly comparable with the standard labour-force measure for the US, but dwarfed by the widely cited non-farm labour-force measure which signalled unemployment in excess of 30 per cent in 1932–3.

The rise in unemployment in Britain and the peak attained were thus far less significant than in the US. However, six issues are particularly important for the British story in comparative context.

  • Unemployment had high political salience even before the 1929 downturn as the 1920s has been widely perceived as a decade of economic underperformance (Pigou’s ‘Britain in the doldrums’ (1947)); the June 1929 election had been dominated by unemployment and an early version of the Keynesian solution (Lloyd George’s Liberal Party claim, ‘We can conquer unemployment’, comprising a £250m 2-year programme of loan-financed public works); and the 1929 downturn was not related to the breaking of a domestic boom.

  • The initial demand shock to the British economy was external, with export-sensitive employment leading at the upper turning-point (Phelps Brown and Shackle, 1939, Diagram 3), whereas in the US depressive forces were internally generated, with employment data (Bernstein, 1987, pp. 155–65) consistent with the Romer (1990) analysis that the ‘great crash’ quickly impacted on consumers’ expenditure and thence employment.

  • Of the then large economies, the UK was the most open, and spectacularly so in relation to the US. The trade openness ratio (exports plus imports/GDP) was only 0.10 for the US but 0.42 for the UK.5 Thus for the UK, but not for the US domestic economy, the international transmission mechanism of depressive impulses and a potential external constraint were essential elements of the macroeconomic circumstances to which policy-makers had to respond.

  • With no breaking of a domestic boom in 1929, there was no asset-price bubble to unwind (this in the context of falling prices with consequent debt-deflation risks) and there was also no domestic banking crisis: thus, relative to the US, the challenge for monetary policy would be different, as would dealing with the issue of business confidence fractured by the fragile real economy.

  • With British central government expenditure at 25.2 per cent of GDP in 1932/3 as against 6.9 per cent for the US federal government in 1932, the demand leverage of the former significantly exceeded that of the latter, with consequently much greater potential for stabilizing demand in Britain than the US if the British (highly centralized economic) authorities so chose (the figures for total public expenditure (both 1932) diverge less, being 27.9 per cent for Britain and 21.2 per cent for the US).6

  • Because of the characteristics of Britain’s fiscal system (a high cyclical macro-marginal budget rate, a measure of the automatic stabilizing properties of Britain’s budgetary system),7 rising unemployment after 1929 quickly translated into a severe budgetary crisis. This peaked in the summer of 1931 and, when combined with a balance-of-payments crisis, resulted in the collapse of the Labour government and the advent of a (Conservative-dominated) National government for the remainder of the decade.

In these circumstances, it is not difficult to understand why mitigating the downturn and seeking to promote recovery became the policy issues that they did for both economic orthodoxy and heterodoxy. However, for a full appreciation of Britain’s situation, both absolutely and in relation to the US, it is necessary to explore more fully macroeconomic performance at a disaggregated level. Thus in Figures 2 and 3 we chart, using annual data, the percentage change in GDP and components for the two economies for three sub-periods expressed as percentages of the 1929 GDP baseline: for the UK, the 1929–32 ‘contraction’ phase, the 1932–7 ‘recovery’, and a composite 1937–9 ‘recession and rearmament’ phase; and for the US, the 1929–33 ‘depression’ phase, the 1933–7 ‘recovery’ phase, and a 1937–41 ‘prewar recession and recovery’ phase. The contrast produces strikingly different results which must inform how we assess the relative policy space and potential efficacy of active fiscal and monetary policies, whether pre-Keynesian or Keynesian-inspired.

Figure 2

UK: change in GDP and components (% of 1929 GDP), 1929–39

Source: Calculated from Sefton and Weale (1995, Table A.3).

Figure 2

UK: change in GDP and components (% of 1929 GDP), 1929–39

Source: Calculated from Sefton and Weale (1995, Table A.3).

Figure 3

US: change in GDP and components (% of 1929 GDP), 1929–41

Source: Calculated from Carter et al. (2006, III, series Ca84-Ca89).

Figure 3

US: change in GDP and components (% of 1929 GDP), 1929–41

Source: Calculated from Carter et al. (2006, III, series Ca84-Ca89).

Starting with the respective contraction/depression phases, we have the scale factor in GDP losses that we have already noted. That said, in many ways it is the compositional story that is more interesting. The UK downturn was most evident in the fall in exports, whereas for the US, export losses were a very small part of the GDP loss, with the dominant depressive impulses the collapse of private consumption and investment expenditures. By contrast, in Britain consumers’ expenditure rose and gross domestic fixed capital formation (GDFCF) only fell slightly during the contraction phase. In both economies, nominal government expenditures rose slightly, but in the US this was dwarfed by the GDP loss. Importantly, for Britain, the contraction phase saw a very marked deterioration in the current-account balance of payments, which slipped from a surplus of 2 per cent of GDP in 1929 to a deficit of 2.4 per cent by 1931 (Middleton, 1996, Figure 7.8), contributing to the severity of the crisis of that year. This has important implications for the Keynesian counterfactual story, and it reaffirms our earlier point about the nascent external constraint confronting policy-makers.

Turning to the recovery phases, there is more similarity between the two national stories but one important difference, which draws our attention to a very probable differential monetary policy affect in Britain (with cheap money and housebuilding) as against the US: this is the failure of GDFCF, which in the US had by 1937 only just reached its 1929 level, whereas in Britain it had attained its 1929 level during 1934 and would by 1937 be nearly 27 per cent higher. We should note also for Britain the neutral affect on GDP of foreign trade during this phase. Lastly, on the final recession and prewar recovery phases, there is much similarity between the two economies with government expenditures both the dominant reflationary impulse. For Britain, this is essentially a rearmament-induced recovery; for the US this is not the case until 1941.

Policy regime and policy-setting

One further stage of scene-setting is necessary if the policies of three generations ago are to be fully comprehended and genuine knowledge transfer made possible. Here we deploy the concept of policy regime. While apparently anachronistic for the interwar period, it is a legitimate device and has been applied usefully by Bordo and Schwartz (1999, p. 151) for monetary policy, as ‘encompass[ing] the constraints or limits imposed by custom, institutions and nature on the ability of the monetary authorities to influence the evolution of macroeconomic aggregates’. The concept of policy regime, of course, relates immediately to the monetary policy trilemma, that the authorities can attain only two out of three possible policy objectives simultaneously: exchange-rate stability; free international capital mobility; and national monetary-policy independence directed towards domestic goals (Obstfeld et al., 2005).

Our account of the British policy regime draws upon Eichengreen (1992) in terms of the key role exercised by the gold standard as a commitment device for generating policy credibility about the implicit policy target of price stability and the explicit goal of maintaining the exchange rate; and, from a public-choice perspective, upon Buchanan and Wagner (1977) on the role played by the nineteenth-century balanced-budget rule in constraining Leviathan. In my own work (principally Middleton (1996)), I have tried to synthesize these approaches to generate an interlocking, nineteenth-century fiscal constitution which was carried forward after the First World War, was fatally disturbed by the events of 1929–31, and then did not survive the Second World War.

Comprising three elements—free trade, the minimal balanced-budget rule, and the gold standard, the traditional defences of the minimalist state—the nineteenth-century fiscal constitution provided the bedrock for British economic policy in the 1920s, and especially after 1925 when the return to the gold standard was secured at the prewar parity. Freedom from political interference for trade, fiscal, and monetary policies was thus the explicit goal; however, it was to be more pronounced as rhetoric than as reality.

  • Free trade had never been complete before 1914 and, with the First World War and its immediate aftermath, acquired a strategic dimension.

  • Fiscal policy was assailed by constant pressure-group lobbying; Churchill as Chancellor used fiscal window-dressing to create the illusion of maintaining nominal balanced budgets, while in any case, the budget was no longer minimal. For the first time outside of war, a British government had real demand leverage and, as we have seen, this was very considerably greater than that available to the US federal government.

  • Monetary policy, both before and during the period of the restored gold standard, saw severe tensions over interest rates (now highly politicized) between Chancellors of the Exchequer and the Governor of the Bank of England, the infamous (and not just in central banking circles) Montagu Norman.

There were important differences in the policy regime as between Britain and the US and not just in relation to the much greater fiscal leverage exercised by the British central government, this a manifestation of both a more highly centralized polity than the US and a significantly different set of preferences for government as against the market. The two countries had very different prewar policy trajectories. Their policy institutions, the economic authorities, and the wider policy communities also differed significantly. British governance was more centralized and the policy community less fragmented than the US (though, in practice, the degree of British central control over public investment was far less than that claimed by those propagating the Keynesian solution; see Middleton (1982)). Above all, if Chancellors were unencumbered by meddling or dissenting prime ministers and other senior Cabinet colleagues, as they largely were between 1929 and 1937 (with first Philip Snowden and then Neville Chamberlain), their authority within the core executive was immense. Thus, at first sight, a determined British government could have a decisive economic impact if it so chose. Of course, and here we return to the Keynesian solution not adopted, it followed a path that was neither that sought by economic progressives nor entirely that for which economic orthodoxy hankered (Middleton, 2004, pp. 477–89).

Trade policies were the most distinctly different, with the dominant narrative being that the Smoot–Harley tariff contributed towards the US downturn becoming a global depression, whereas, for the UK, the formal abandonment of free trade (a temporary measure in November 1931, codified in April 1932) is part of the standard explanatory set—albeit contested—for Britain’s domestic economic recovery from 1932. For fiscal policy, there is obviously considerable overlap on the issue of the Keynesian solution, but it is important not to lose sight of the pressures for public-sector growth which were independent of the stabilization debate and were stronger in the UK than the US.

On monetary policy, the Federal Reserve System was new relative to the Bank of England and very differently structured, with—post Benjamin Strong—the US less committed to promoting international economic cooperation. There was no banking crisis in Britain, no burst asset-price bubble, whereas in the US it would be impossible to explain monetary policy without starting with bank failures. Moreover, the very different economic experiences of the war impinged on the policy space, not least in terms of domestic and international indebtedness. Figure 4 charts the transformation of three key measures of debt dynamics (the ratios of government debt to GDP, and debt interest to total public expenditure and to GDP), with the Second World War and immediate postwar period shown, as well as the Great War, to illustrate the relatively greater fiscal significance of that first conflict. Given the significant price deflation between 1920 and 1935 it is not difficult to see why debt management was to dominate both monetary and fiscal policy (Middleton, 1985, ch. 6; Foreman-Peck, 2002). Moreover, Figure 4 charts the combined public sector; for the central government budget, which was the focus of attention, debt-service payments (which on contemporary conventions required a significant sinking fund) were appropriating at least 40 per cent of government expenditure from the early 1920s until 1930/1, at which point, in extremis, first, the sinking fund was suspended and then cheap money delivered its substantial fiscal dividend (Middleton, 1985, Table 6.4).

Figure 4

UK: outstanding national debt, debt interest and total public expenditure (% of GDP), 1913–50

Sources: National debt outstanding at end of financial year: Mitchell (1988, pp. 602–3); debt interest, total public expenditure, and GDP: Middleton (1996, Table AI.1).

Figure 4

UK: outstanding national debt, debt interest and total public expenditure (% of GDP), 1913–50

Sources: National debt outstanding at end of financial year: Mitchell (1988, pp. 602–3); debt interest, total public expenditure, and GDP: Middleton (1996, Table AI.1).

Finally, exit from the gold standard was very different for the two countries: the US acted unilaterally in 1933, whereas Britain was forced to abandon gold in September 1931 as a toxic mix of budgetary and balance-of-payments crises morphed into a political crisis which made the status quo unsustainable. In the following 9 months there then occurred a reorientation of economic policy that impacted on each element of the monetary policy trilemma. It is this policy reorientation which is the dividing line for interwar British economic policy: between a macroeconomic policy passivity born of the gold standard constraint and a post gold-standard activism (albeit non-Keynesian) when freed from the ‘golden fetter’. Moreover, as Eichengreen and Sachs (1985) have argued in what has become the orthodoxy, almost a stylized fact of the 1930s: those that left the gold standard early tended to be the first to recover in the 1930s. That said, a basic methodological problem has bedevilled all work on the impact of interwar macroeconomic policy, and especially post-September 1931 when sterling depreciation and policy reorientation brought, in sequence, a managed floating exchange rate; tariff protection for manufactures; ‘cheap money’ and regained monetary-policy independence; and, after a longer lag, even a loosening of the fiscal stance. With a certain understatement, Solomou (1996, p. 112) writes that such was the clustering of these momentous policy changes that it is ‘extremely difficult to distinguish individual policy impacts’.

IV. Policy: narrative and assessment

After conducting a thorough literature review,8Worswick (1984, p. 93) concluded emphatically: ‘Spontaneous forces alone will not count for the recovery [of 1932–7]. Policy made important contributions.’ Discounting a ‘natural’ recovery, and in particular following Dimsdale (1984, p. 102) that the fall in employment during the contraction phase and its recovery thereafter were largely accounted for on the demand side, we here follow Worswick who, in identifying the proximate cause of the depression as a sharp decline in exports, quite naturally accorded the abandonment of gold and the ensuing depreciation of sterling as at minimum permissive of recovery. In the recovery phase, net export growth made no contribution to GDP growth (Figure 2), and we are thus looking to explain a recovery in domestic demand in terms of domestic policy.

Figures 5–8 provide the background for our policy narrative and the starting point for the assessment of policy effectiveness. Figure 5 charts monthly movements in the two official interest rates, the Bank of England base rate and—since London perceived itself as a price taker during the gold standard era—the New York Federal Reserve discount rate; Figure 6 consumer prices, both as index numbers (1914 = 100) and as percentage change on the previous year; Figure 7 the nominal $–£ and three effective exchange-rate measures; and Figure 8 public-sector budget balances, this for a much longer period than Figures 5–7 to illustrate that there is much misunderstanding about the public finances and thus budgetary orthodoxy in the pre-Keynesian era. The combined balance, of current plus capital accounts, was rarely in surplus—for 5 years only between the wars (1920–38), and interestingly two of these were 1933 and 1934, this a testament to the fiscal tightening of 1929–33. However, the current balance was positive in 15 out of the 18 years, while the primary balance (an important indicator in today’s debate about debt dynamics and fiscal consolidation) was positive every year, averaging 7.4 per cent of GDP for the period.

Figure 5

Official interest rates, monthly averages: UK and US, 1929–40

Sources: UK: http://www.bankofengland.co.uk/statistics/rates/baserate.xls, 08.04.10; US: http://www.nber.org/ databases/macrohistory/contents/chapter13.html, 08.04.10.

Figure 5

Official interest rates, monthly averages: UK and US, 1929–40

Sources: UK: http://www.bankofengland.co.uk/statistics/rates/baserate.xls, 08.04.10; US: http://www.nber.org/ databases/macrohistory/contents/chapter13.html, 08.04.10.

Figure 6

Consumer price index (1914 = 100), UK and US, 1914–40

Source: Derived from Maddison (1991, Tables E2-4).

Figure 6

Consumer price index (1914 = 100), UK and US, 1914–40

Source: Derived from Maddison (1991, Tables E2-4).

Figure 7

Sterling exchange rates (1929 = 100), 1929–38

Sources: Redmond (1980); Dimsdale (1981, Tables 3, 9).

Figure 7

Sterling exchange rates (1929 = 100), 1929–38

Sources: Redmond (1980); Dimsdale (1981, Tables 3, 9).

Figure 8

Public-sector budget balances, % of GDP, 1900–39

Source: Derived from Middleton (1996, Tables AI.1–AI.2).

Figure 8

Public-sector budget balances, % of GDP, 1900–39

Source: Derived from Middleton (1996, Tables AI.1–AI.2).

Monetary policy

Monetary policy divides relatively neatly at September 1931 when Britain was forced to abandon the gold standard, although two important considerations pertain in terms of policy space and the policy record. First, although pre-September 1931 monetary policy was often tighter than was desirable in light of the developing downturn, it is clear from the behaviour of the monetary aggregates and the policy record that the authorities were ‘more focused on domestic objectives than on following the “rules of the [gold standard] game”’ (Bernanke, 2000, pp. 151–2), lowering Bank Rate (from 6 to 2.5 per cent between 1929Q3 and 1931Q2) and sterilizing the effects of gold outflows on the money supply (Howson, 1975, pp. 43–4, 66–9; Bernanke, 2000, Table 4.c). Second, the abandonment of gold coincided with a temporary, substantial increase in bank rate (from 4.5 to 6 per cent); it was to take some months (an anxiety phase) before the authorities felt sufficiently confident to contemplate lower interest rates, a process greatly aided by tightening capital controls and financial turbulence elsewhere, which made London appear a relatively safe heaven for highly mobile, short-term capital. Thus the end of the golden fetter did not bring an immediate freedom of action, though the potential for reorientation was recognized from the outset (Howson, 1975, pp. 79–89).9

For the beginning of our survey period there is a broad consensus: that ‘money played at best only a minor role in the economic depression of 1929–32’; and that, by abandoning gold, Britain ‘dodged’ the monetary collapse experienced by the US (Capie and Wood, 1994, pp. 236, 242). Monetary-policy independence, of course, did not exist with the gold standard. However, once a floating—indeed, highly managed—exchange rate was secured and policy-makers had recovered from the initial panic about inflation (fears that an uncontrollable depreciation of sterling would result in imported inflation, this being an added reason for quickly securing fiscal retrenchment to maintain confidence), Bank rate was reduced significantly (to 2 per cent) and was maintained thereafter (bar a short-lived spike associated with the onset of the war).

The standard questions asked about the role of monetary policy after September 1931 concern its effectiveness with respect to:

  • exchange-rate objectives;

  • the promotion of economic recovery through higher investment, and especially in the housing market;

  • the relaxation of the fiscal stance made possible by the significant reduction in the interest burden of the national debt; and

  • the extent to which there was a trade-off between cheap money and greater fiscal activism.

Taking each in turn, beginning with exchange-rate policy: this, after the anxiety phase of autumn–winter 1931, quickly and quietly evolved into a managed float of sterling designed inter alia to give some competitive edge to British exports without undue stress upon the geo-politics of international finance and of the British empire (Howson, 1980).

On the second, the key question of the cheap-money era, there have been a number of studies of the sensitivity of the components of interwar aggregate demand to interest rates. Broadberry’s (1986, pp. 64–5, 129, 131, 142–4) aggregate demand–supply model for interwar Britain, which has a monetary dimension though not a formal monetary sector, is widely cited on the effects of cheap money. He found a low interest elasticity for investment (–0.11 for non-housing and (long-run) –1.1 for housing) with the major effect of monetary policy transmitted via the external sector through the exchange rate, contributing to the 13 per cent depreciation of sterling between 1931 and 1933 which raised output by about 3 per cent through the improvement in the balance of trade. Much, of course, has been claimed for the role of a house-building boom in the recovery of the 1930s, along with an older strand of structural change in industry (the ‘old–new’ industries debate) partly driven by rising real incomes over the depression period. Broadberry (1986, pp. 25, 61) allocates about half the rise in housing investment in 1931–3 to the advent of cheap money, with net investment in housing exceeding net total investment in 1932–3. These broad parameters, of cheap money being most effective in the housing sector, are confirmed by Dimsdale and Horsewood (1995).

Similarly, from their review of the monetary data, Capie and Wood (1994, p. 246) confirmed that cheap money ‘undoubtedly encouraged investment’, and that while ‘money was not central . . . to the recovery . . . doubtless it was accommodating’, though it should be noted that this, of course, is from the contested perspective that, in the long run, monetary conditions do not determine real interest rates (cf. Chadha and Dimsdale, 1999). We have then a positive role for cheap money (Dimsdale, 1984; Worswick, 1984): in terms of timing (coincident with the lower turning-point in 1932) and significance, it impacted strongly on residential investment and supported the recovery of the industrial and commercial sector. While money-supply growth was not continuous over the recovery phase, relative to European competitors (many still fettered by the gold standard), it was undoubtedly helpful for recovery and this notwithstanding the Treasury’s ‘funding complex’ with respect to the floating debt which was in conflict with the cheap money policy, thereby lessening the fall in long-term interest rates (Howson, 1975, pp. 95–103). The contrast with the US is twofold: first, the absence of banking crises and, second, that in Britain the positive contribution of policy to the recovery was distributed between monetary and fiscal policy, whereas in the US little is claimed at best for fiscal policy and it is the monetary expansion that is highlighted (Romer, 1992).

As regards the third question, it is arguable that further tax rises and/or retrenchment in expenditure were avoided in 1932/3 and above all 1933/4 because of the lessened debt charge on the budget: the reduction in total debt service including sinking funds between 1932/3 and 1933/4 was nearly 2.5 percentage points of GDP (see Figure 4). Without cheap money the recovery could well have stalled.

The fourth question, the trade-off, we defer until our treatment of fiscal policy and the Treasury view of the (in)efficacy of long-financed public works (see section IV(iii)). At this stage, all we need to report is that key policy-makers in the Treasury and Bank of England were certain that cheap money could be imperilled by an inappropriate setting of fiscal instruments.

Fiscal policy

Overall position

Fiscal policy is both simple and incredibly complex to evaluate between the wars: apparently simple to comprehend, because the goals appear straightforward, being, as pre-war, the observance of the minimal balanced budget rule; in practice, however, incredibly complicated, as the period witnessed a complex of enormous pressures for expenditure growth (these additional to the Keynesian stabilization debate), a budget now significantly enlarged on pre-war and highly cyclically sensitive, all of this within the context of considerable macroeconomic instability and the pressures for the Keynesian solution, which, importantly, predate the 1929 downturn.

The traditional view of fiscal policy in the 1930s is one of a rigid pursuit of orthodoxy (Richardson, 1967, pp. 211–12), whereby the authorities sought an ex post budget balance and, if ex ante this was threatened by a macroeconomic shock, they were prepared to override the automatic stabilizers to regain budgetary equilibrium. On this view, fiscal orthodoxy equated to inherent macroeconomic policy destabilization. Policy-makers set the budget, and the fiscal stance was assessed, in terms of the actual budget balance (for central, not general, government). As a summary measure of fiscal influence this is, of course, subject to the well-known problem, explored first for the US in Brown (1956), of the endogenity of the budget. Middleton (1981), whose results are in Figure 9, first adjusted the actual central government budget balance (Bc/Y) for the authorities’ fiscal window-dressing to produce a new balance measure (B/Y) and then used a Brown (1956)-type methodology to calculate a constant employment surplus measure (B*/Y*).10 On this basis, fiscal policy was contractionary throughout the contraction phase and well beyond, though it did—via rearmament—impart a very significant stimulus which ensured that the 1937–8 recession was of short duration and that the economy was growing rapidly by the eve of the Second World War. Indeed, Thomas (1983, p. 571) argues: ‘The success of rearmament in creating employment, even at the top of the cycle, leads us to view the eschewment of fiscal policy in the thirties as a missed opportunity for the economy.’

Figure 9

Summary measures of fiscal stance, % of actual and constant employment GDP, 1929/30–1939/40

Source: Middleton (1981, Table 5).

Figure 9

Summary measures of fiscal stance, % of actual and constant employment GDP, 1929/30–1939/40

Source: Middleton (1981, Table 5).

The overall position can be demonstrated in a different fashion, conceptualized in Figure 10 to highlight the issue of scale (the smallness of the deficits in relation to the output gap) and to enable comparison with Fishback’s (2010, Figure 4) identical presentation for US fiscal policy. The mildness of the output loss relative to the US is again immediately apparent (mirroring what we have demonstrated in Figures 1–3), but what is new is the comparative stability of total public expenditure and total revenue (both here measured at actual employment and for the combined public sector as against the central government accounts used for the fiscal stance debate discussed above) and especially in relation to the growing output gap to 1933 which amounted to over £510m (nearly 11 per cent of 1929 GDP). Against this contraction, the public-sector deficit was reduced by £50m between 1929 and 1933, although over 1929–31—the period in which the authorities attempted, but did not succeed fully, in overriding the automatic stabilizers—the actual deficit did widen. Nonetheless, this was only by 16 per cent of the 1931 output gap.

Figure 10

GDP deviations, total public expenditure, and total revenue, 1938 market prices, 1929–39

Sources: GDP: Feinstein (1972, Table 5); TPE and TR: Middleton (1996, Tables AI.1, AI.2) adjusted by GDP deflator, Feinstein (1972, Table 61).

Figure 10

GDP deviations, total public expenditure, and total revenue, 1938 market prices, 1929–39

Sources: GDP: Feinstein (1972, Table 5); TPE and TR: Middleton (1996, Tables AI.1, AI.2) adjusted by GDP deflator, Feinstein (1972, Table 61).

Finally, and again on the basis of data not cyclically adjusted, we can decompose both the expenditure and revenue sides of the combined public sector (Table 1) to derive the following results about the path of fiscal policy. For the contraction phase, we note first that the combined public authorities’ deficit fell from 0.7 to 0.5 per cent of GDP between 1929 and 1932; the product of an increase in the expenditure ratio of 3.4 percentage points of GDP and the receipts ratio by 3.6 percentage points. For expenditures which could be considered to be exhaustive or nearly so on the first round (current goods and services plus gross capital formation plus current grants to the personal sector) there was a rise of 3.3 percentage points, but against this must be set tax rises of 3.6 percentage points, with taxes on income rising slightly more than taxes on expenditure. The prevailing ethos among the interwar business community was that taxes on income were the most deflationary, and the US example suggests that further research is now needed on the supply-side effects of fiscal policy: how tax rates and tax structures impacted upon spending and investment. Looking at the recovery phase, 1932–7, the receipts ratio was reduced by five times more than the expenditure ratio, showing how difficult was fiscal consolidation on the expenditure side and how, with the recovery, taxes on income, in particular, could be relaxed. The beginnings of rearmament are also evident on the expenditure side. It is noteworthy that at the end of this phase there was a deficit of 1.5 per cent of GDP, this in the circumstances of a cyclical peak, whereas in the depth of the crisis the 1931 deficit had only been 2.3 per cent of GDP. Finally, the rearmament phase sees an unprecendented peacetime surge in expenditure on current goods and services, this financed by borrowing as the combined authorities’ deficit was by 1939 some 8.3 per cent of GDP, making the fiscal stimulus of 1937–9 the most pronounced experienced by the British economy in peacetime.

Table 1

Public-sector accounts by economic classification, changes in % points of GDP at actual employment, 1929–39

A. Total public expenditure 
 Current goods and services Gross capital formation Current grants to personal sector Subsidies Current grants paid abroad Debt interest  Total 
1929 9.2 2.6 4.4 0.5 0.1 7.7  24.5 
1930 9.5 2.8 5.1 0.5 0.1 7.6  25.5 
1931 10.2 3.3 6.5 0.5 0.1 7.7  28.2 
1932 10.1 2.8 6.6 0.6 0.1 7.8  27.9 
1933 10.1 2.2 6.4 0.7 0.1 7.0  26.5 
1934 9.9 2.1 5.9 0.7 0.1 6.2  25.1 
1935 10.2 2.4 5.8 0.8 0.1 6.0  25.3 
1936 10.9 2.9 5.4 0.7 0.1 5.7  25.7 
1937 11.7 3.3 5.0 0.6 0.1 5.4  26.0 
1938 13.4 3.6 5.0 0.7 0.1 5.2  28.1 
1939 19.8 2.9 4.5 0.8 0.3 5.0  33.2 
Change (% points of GDP): 
 1929–32 0.9 0.2 2.2 0.1 0.0 0.1  3.4 
 1932–7 1.6 0.5 –1.6 0.0 0.0 –2.4  –1.9 
 1937–9 8.1 –0.4 –0.5 0.2 0.1 –0.4  7.1 
 
B. Total receipts 
 Taxes on income Taxes on expenditure Taxes on capital National insurance contributions Gross trading surplus Rent, interest and dividends Current grants from abroad Total 
1929 6.2 10.6 1.7 1.7 1.1 2.0 0.5 23.8 
1930 6.5 10.3 1.7 1.7 1.2 2.2 0.6 24.1 
1931 7.5 11.0 1.7 1.9 1.3 2.1 0.5 25.9 
1932 8.2 12.0 1.7 2.1 1.4 1.9 0.0 27.4 
1933 7.3 12.1 2.0 2.1 1.5 1.9 0.0 26.9 
1934 6.4 12.0 1.7 2.1 1.5 1.8 0.0 25.6 
1935 6.0 11.8 1.8 2.1 1.4 1.8 0.0 25.0 
1936 5.9 12.0 1.8 2.1 1.4 1.8 0.0 25.0 
1937 6.2 11.6 1.8 2.0 1.3 1.7 0.0 24.5 
1938 6.9 11.3 1.4 2.0 1.3 1.7 0.0 24.4 
1939 7.4 11.5 1.3 1.8 1.3 1.6 0.0 24.9 
Change (% points of GDP): 
 1929–32 2.0 1.5 0.0 0.3 0.3 –0.1 –0.5 3.6 
 1932–7 –2.1 –0.5 0.1 –0.1 0.0 –0.3 0.0 –2.9 
 1937–9 1.2 –0.1 –0.5 –0.2 –0.1 –0.1 0.0 0.3 
A. Total public expenditure 
 Current goods and services Gross capital formation Current grants to personal sector Subsidies Current grants paid abroad Debt interest  Total 
1929 9.2 2.6 4.4 0.5 0.1 7.7  24.5 
1930 9.5 2.8 5.1 0.5 0.1 7.6  25.5 
1931 10.2 3.3 6.5 0.5 0.1 7.7  28.2 
1932 10.1 2.8 6.6 0.6 0.1 7.8  27.9 
1933 10.1 2.2 6.4 0.7 0.1 7.0  26.5 
1934 9.9 2.1 5.9 0.7 0.1 6.2  25.1 
1935 10.2 2.4 5.8 0.8 0.1 6.0  25.3 
1936 10.9 2.9 5.4 0.7 0.1 5.7  25.7 
1937 11.7 3.3 5.0 0.6 0.1 5.4  26.0 
1938 13.4 3.6 5.0 0.7 0.1 5.2  28.1 
1939 19.8 2.9 4.5 0.8 0.3 5.0  33.2 
Change (% points of GDP): 
 1929–32 0.9 0.2 2.2 0.1 0.0 0.1  3.4 
 1932–7 1.6 0.5 –1.6 0.0 0.0 –2.4  –1.9 
 1937–9 8.1 –0.4 –0.5 0.2 0.1 –0.4  7.1 
 
B. Total receipts 
 Taxes on income Taxes on expenditure Taxes on capital National insurance contributions Gross trading surplus Rent, interest and dividends Current grants from abroad Total 
1929 6.2 10.6 1.7 1.7 1.1 2.0 0.5 23.8 
1930 6.5 10.3 1.7 1.7 1.2 2.2 0.6 24.1 
1931 7.5 11.0 1.7 1.9 1.3 2.1 0.5 25.9 
1932 8.2 12.0 1.7 2.1 1.4 1.9 0.0 27.4 
1933 7.3 12.1 2.0 2.1 1.5 1.9 0.0 26.9 
1934 6.4 12.0 1.7 2.1 1.5 1.8 0.0 25.6 
1935 6.0 11.8 1.8 2.1 1.4 1.8 0.0 25.0 
1936 5.9 12.0 1.8 2.1 1.4 1.8 0.0 25.0 
1937 6.2 11.6 1.8 2.0 1.3 1.7 0.0 24.5 
1938 6.9 11.3 1.4 2.0 1.3 1.7 0.0 24.4 
1939 7.4 11.5 1.3 1.8 1.3 1.6 0.0 24.9 
Change (% points of GDP): 
 1929–32 2.0 1.5 0.0 0.3 0.3 –0.1 –0.5 3.6 
 1932–7 –2.1 –0.5 0.1 –0.1 0.0 –0.3 0.0 –2.9 
 1937–9 1.2 –0.1 –0.5 –0.2 –0.1 –0.1 0.0 0.3 

Source: Derived from Middleton (1996, app. I).

Fiscal policy and fiscal consolidation

Those seeking lessons from the early 1930s for the present will find very instructive what can happen when the authorities override the automatic stabilizers. To pursue this, we have first to ground policy in terms of contemporary parameters. These were that it was the actual budget balance (Bc) which was the focus of contemporaries’ attention and that, by earlier wartime and later post-1945 standards, the ex post deficits were very small in relation to GDP.

For the pre-crisis year of 1930/1, the authorities were facing a prospective deficit of £47.3m (on expenditure of £870.9m) but after raising taxation and other measures in the April 1930 budget, in effect over-riding the automatic stabilizers, they budgeted for a small surplus of £2.2m. In the event, 1930/1 closed with a deficit of £23.2m (Middleton, 1985, Table 6.2), but this was only 0.6 per cent of GDP. In the following year, the crisis year, the May committee report published on 31 July 1931 forecast a deficit for 1932/3 of £120m of expenditure over income, with this becoming the focus for what became the crisis that on 24 August brought down Labour and established a National government (the gold standard fell less than a month later). A prospective deficit of £120m was viewed widely as enormous, requiring urgent action lest there be fatal injury to confidence and thence to the gold standard (Williamson, 1992, pp. 267–73).11 A deficit of £120m represented 3.1 per cent of 1931/2 GDP, a magnitude which would become routine during the Keynesian era of the 1950s and 1960s, and, of course, in a year of extreme stress is small relative to deficits in Britain since 2007. Additionally, in 1930/1 and 1931/2 the balance of the British budget was closer to equilibrium than in the US, France, and Germany (Middleton, 1985, Table 6.8). It was Britain’s special position, and the perception at home and abroad of heavy British indebtedness, that dictated that it adhere more strictly to orthodox financial principles than was considered to be necessary elsewhere.

The emergency budget of September 1931 was the turning point for fiscal policy that decade. This made operational a somewhat watered down version of the May report; indeed, it bore a close resemblance to that which the outgoing Labour government had agreed but which had been deemed insufficient by the financial community, domestic and international. The May committee had proposed a fiscal consolidation weighted at 80 per cent retrenchment (of which two-thirds would be cuts in unemployment benefits) and 20 per cent additional taxation, amounting to £120m. The new government’s budget made adjustments totalling £76m for 1931/2, comprising additional taxation (£40.5m), retrenchment (£22m), and reduction in the sinking fund (£13.7m) (Middleton, 1985, Table 6.2). Even if we count the sinking fund as expenditure, this gives fiscal consolidation which was now biased towards additional taxation (53 per cent); it also demonstrates that the National, unlike the Labour, government was able to operate with greater latitude in terms of its adherence to fiscal conventions (the sinking fund) and the extent of consolidation that would satisfy nervous markets.

For economists, the story here is of a tightening of the fiscal stance amid a depression known to be deepening, but if we dig deeper into the mechanics of budgetary policy and what we know about policy-makers’ thinking at this time, we arrive at a more nuanced picture. This was put very well by an historian who had no access to the offical papers (they were not released until the 1970s) but who could read the runes:

The deflationary phase of the National government’s policy was short-lived . . . [their] financial policies made the best of both worlds; they seemed sufficiently deflationary to restore confidence; they were in fact sufficiently inflationary to assist recovery by maintaining the purchasing power of the people. (Mowat, 1955, p. 455)

Chancellors and their senior officials understood full well the undesirability of raising taxes in a depression; at least some appreciated the likely net adverse impact on purchasing power of cutting transfer payments; and a few may even have accepted the Keynesian logic that discretionary action to balance the budget by overriding the automatic stabilizers risked reducing GDP and thus making budgetary equilibrium more difficult to attain in practice, something Keynes (1931, pp. 144–5) articulated for the first time convincingly during this crisis. Whatever their understanding, the reality for the policy space was that orthodoxy must be seen to prevail. That, covertly, the British Treasury had for some time been engaging in fiscal window-dressing (at times amounting to 1 per cent of GDP) to improve the ex post budget balance (Bc) suggests that it understood the dangers of overriding the automatic stabilizers and that this device served the purpose of lessening the extent of fiscal tightening required in a depression to maintain a balanced budget.

The macroeconomic effects of the 1930–2 fiscal tightening have not been much examined in detail. In particular, we know little about what we might call the pass-through effects for the real economy of implementing the autumn 1931 fiscal consolidation.12 The package adopted slightly favoured higher taxation over expenditure retrenchment (a ratio of 53 to 47 per cent); and of the tax rises, it was taxes on income as against expenditure which bore the main brunt, including a 2.5 points rise in the standard rate of income tax (to 25 per cent) and a 10 per cent surcharge on the supertax (which gave progression to the income tax system), though there were additional reliefs for companies (there was no separate corporation tax at this time). The economic consequences of these tax changes are difficult to establish, and especially when income tax was actually paid in arrears (up to 2 years for the self-employed and for companies’ profits). Nonetheless, we would, ceteris paribus, expect a depressive effect on incomes, spending, and thus employment, especially when we factor in the background of trend falling prices which impacted to raise the effective rate of the majority of taxes on expenditure, this the most important element of the tax base.

The likelihood of depressive effects becomes stronger as we examine the expenditure side: of the £70m of cuts planned for 1932/3, at least 15 per cent were planned public-sector wage/salary cuts, a maximum of 34 per cent from cuts in expenditure on goods and services (current and capital), and 51 per cent from cuts in the net cost of the unemployment insurance scheme (combined reduced eligibility and benefit levels). The potential impact on consumer spending from salary and benefit cuts is relatively straightforward when such a high proportion of those affected were credit-constrained, but impossible to determine with precision when we lack key quarterly/monthly data. That said, we know that consumers’ expenditure, having grown in the first 2 years of the contraction phase, now fell in 1932, but only by 0.6 per cent. We know further that a gross reduction in the purchasing power of the insured unemployed and employed of some £36m at maximum was not even 1 per cent of 1931 consumers’ expenditure (it is also, of course, dwarfed by the £100m per annum public works programme proposed in 1929, and especially if multiplier effects are taken into account). Moreover, the seasonally adjusted quarterly activity indices we have for this period all have industrial production as fluctuating but broadly trendless between 1931Q3–1933Q1, though employment and quarterly GDP data are less clear-cut and might suggest a recovery in mid-1931 which was temporarily halted and did not resume until 1932Q2 or Q3 (Capie and Collins, 1983, Tables 1.16, 1.36, 4.9; Mitchell et al., 2009, Tables 1A, 2A). Even so, the key point is, however big the fiscal consolidation was as a political event, it was far less so in terms of macroeconomic impact. Furthermore, in relation to the modern literature on achieving successful fiscal reform (OECD, 2007), which emphasizes permanent expenditure cuts over increased taxation, the politics of the 1930s coalition government precluded what we now think to be the more robust consolidation path.

Table 2

Simulated effects of a £100m public-spending programme, 1930

Multiplier value Low Medium High 
(1.25) (1.50) (1.75) 
A. With fixed exchange rate: 
 Income change (£ millions) 125.0 150.0 175.0 
 Employment change ('000s) 345.6 414.8 483.9 
 Change in budget balance (£ millions) –50.0 –40.0 –30.0 
 Change in current account –26.3 –31.5 –36.8 
  balance of payments (£ millions)    
B. With floating exchange rate: 
 Income change (£ millions) 167.0 214.0 269.0 
 Employment change ('000s) 461.8 591.7 743.8 
 Change in budget balance (£ millions) –33.2 –14.4 7.6 
 Exchange rate change (%) –5.1 –6.5 –8.2 
Multiplier value Low Medium High 
(1.25) (1.50) (1.75) 
A. With fixed exchange rate: 
 Income change (£ millions) 125.0 150.0 175.0 
 Employment change ('000s) 345.6 414.8 483.9 
 Change in budget balance (£ millions) –50.0 –40.0 –30.0 
 Change in current account –26.3 –31.5 –36.8 
  balance of payments (£ millions)    
B. With floating exchange rate: 
 Income change (£ millions) 167.0 214.0 269.0 
 Employment change ('000s) 461.8 591.7 743.8 
 Change in budget balance (£ millions) –33.2 –14.4 7.6 
 Exchange rate change (%) –5.1 –6.5 –8.2 

Source: Hatton (1987, Table 7.2).

The remainder of the fiscal policy narrative can be quickly recounted. The setting of the budget for 1932/3 was something of a gamble. Had the recovery not commenced, cheap money not been successfully implemented, and fiscal window-dressing not deployed extensively, then the year would almost certainly have closed with a substantial deficit. As it was, a deficit of £32.3m (0.9 per cent of GDP) was posted, but this was not corrosive of confidence in the way that it would have been even 12 months earlier, and by the following year a surplus of almost the same sum was achieved. Thereafter, for the next 3 years, recovery in the public finances tracked recovery in the real economy, providing scope thereby for tax remissions and a resumption of expenditure growth. There was then a brief interruption because of the 1937–8 recession, this like 1929 largely externally initiated, but more importantly, once the decision was made in 1937 to finance by loan a significant part of the rearmament programme, British budgetary policy became as much directed by military as civilian priorities. That said, a number of historians of policy have argued that within the management of the rearmament programme one can detect evidence of the economic authorities employing the logic of Keyensian budgetary policy to manage aggregate demand and taxable capacity (Howson and Winch, 1977, pp. 144–52; Middleton, 1985, pp. 118–19; 173). Of course, the exigencies of rearmament were considered a wholly different justification by contemporaries for deficit-financing. The Keynenian solution thus still had no traction as a peacetime device.

The Keynesian solution

The Great Recession has revived interest in another key aspect of Keynesian stabilization policies: those of using loan-financed public works to mitigate unemployment; what we would now call infrastructure-led growth. This is of twofold interest: first, the debate itself for what it reveals about policy-makers’ conception of what could be achieved and over what timescale, given the dominance of orthodox financial opinion in the context of grave anxieties about new debt creation; and second, because the empirical literature of itself reveals much about the potency of fiscal policy had the Keynesian path been taken, whether this had involved being more accommodating of the automatic stabilizers, pursuing a big public works programme, or, as was suggested by Keynes in The Means to Prosperity (Keynes, 1933), in combining public works with a tax cut. Here we do not concentrate on the debate per se but instead on quantitative assessment of fiscal simulations and their likely impact.

That the Keynesian solution was not adopted in the 1930s does not mean that public works were not pursued, at least in part for employment purposes, or that public-sector investment was not an important component of aggregate demand. Public works on a significant scale had been used just after the First World War and again by the Labour government of 1929–31. In the latter case, programmes amounting to £185m had been approved by June 1931 (many programmes would soon be suspended or scaled back as part of the May committee cuts), of which £108m were in operation. Direct employment creation has been estimated at approximately 100,000 man-years for that month (Middleton, 1983, p. 361), at which point total insured unemployment was 2.74m or 21.2 per cent.

Public-sector investment, which was dominated by local spending and thus was far from the immediate control of central government, averaged 2.7 per cent of GDP between 1920 and 1938. It was marginally higher during the contraction phase (2.9 per cent), a reflection of the Labour government’s 1929–31 public works programme, but then fell back to an average of 2.6 per cent over the recovery phase (Middleton, 1996, Table AI.1). However, on the issue of stabilizing effectiveness, and for the 1930s as a whole, deviations in public investment from trend were neither synchronized nor equivalent in scale to those of deviations in private investment. Over the whole interwar period public investment was destabilizing (Middleton, 1985, pp. 46–7).

This leads inextricably to the wider issues of the potential efficacy of a fiscal stimulus, upon which there have been a number of econometric exercises that have simulated the counterfactual Keynesian solution. Again, there is a rich literature, but here we concentrate on the two exercises from which generalizable conclusions are most obtainable and/or embody the latest data and econometric techniques. First, Hatton (1987) simulated a partial ‘We can conquer unemployment’ package of £100m (equivalent to 2.5 per cent of GDP) in 1930 on two exchange-rate scenarios and with a range of values for the short-run multiplier. The results are summarized in Table 2. Unsurprisingly, the income and employment effects were greatest if exchange-rate regime flexibility was permitted, but that was contrary to policy and even so the simulated effects would not support Lloyd George’s claim, made for the 1929 general election and with Keynes’s backing, that a public-works programme of £100m would bring unemployment back to ‘normal’ levels within a year. Given the fiscal policy multipler, the proposed programme was simply too small in relation to the problem. Subsequently, on the basis of research on the supply side of the interwar economy, much of which he conducted, Hatton has become more pessimistic about the likely effectiveness of the Keynesian solution because of labour-market inflexibility, though unlike the US, where this was exacerbated by policy in the 1930s (the New Deal), resilience was basically the same in Britain in the 1930s as the 1920s (Hatton and Thomas, 2010).

By contrast, Dimsdale and Horsewood (1995) are more optimistic about the multiplier and thus Keynesian solution effectiveness, and this from the perspective of a modelling exercise which now incorporates a developed supply side. Table 3 summarizes their main simulations, principally 3- and 10-year policy packages (of £100m per annum spending programmes), both with and without crowding-out. In relation to the Keynes and Henderson’s (1929) claim that £100m per annum sustained for 3 years would yield an increase in employment of 500,000, they generate much lower results: 333,000 on simulation 1, and 303,000 on simulation 3. However, given the low elasticity of employment, the GDP affect is more pronounced after 3 years: respectively, rises of 4.9 and 4.4 per cent depending upon whether crowding-out is deemed a factor. Prices are also higher, as are real wages but the current-account balance of payments has significantly weakened and this notwithstanding the assumption of a flexible exchange rate. If the policy package is, instead, sustained over 10 years to 1938 (simulations 2 and 4), the increase in GDP approaches 5 per cent and gradually rises thereafter, implying a long-run multiplier of about 2.5 and again with the impact of the Keynesian solution stronger on output than employment. However, we should note the results for the GDP deflator and the balance of payments. With the average current-account balance between 1929 and 1938 of –0.3 per cent of GDP (Feinstein, 1972, Tables 3, 15), these simulations certainly raise questions about the sustainability of such counterfactual spending packages. Dimsdale and Horsewood are clear that the output and employment rises ‘would not have been undermined by crowding-out effects, because the feedback of a rise in interest rates turns out to be relatively weak’ and that the Treasury view ‘does not turn out to be a major obstacle to the effectiveness of the Henderson–Keynes proposals’, but they do accept that ‘the impact of crowding-out effects is somewhat greater when fiscal expansion is maintained over a decade’ (p. 391).

Table 3

Simulated effects of £100m spending programmes, 1929–38

Simulation Description Year Real GDP (%) Employment ('000s) GDP deflator Real wages Current balance (£m) 
£100m p.a. for 1929–31, constant short-term interest rates 3.40 120.3 1.01 0.34 0.29 
4.46 190.5 2.34 0.44 –11.30 
4.87 332.7 3.95 0.65 –47.51 
1.53 270.8 5.40 0.72 –50.62 
0.26 –7.4 7.36 0.74 –52.40 
£100m p.a. for 1929–38, constant short-term interest rates 3.40 120.3 1.01 0.34 0.29 
4.46 190.5 2.34 0.44 –11.30 
4.87 332.7 3.95 0.65 –47.51 
4.98 403.4 5.95 0.90 –50.54 
5.68 321.6 24.45 2.65 –170.03 
£100m p.a. for 1929–31 with crowding-out effects 3.13 110.7 0.93 0.31 0.27 
4.16 177.3 2.15 0.40 –10.30 
4.39 302.5 3.61 0.60 –47.66 
1.42 249.7 4.94 0.66 –49.30 
0.43 9.2 7.64 0.78 –54.53 
£100m p.a. for 1929–38 with crowding-out effects 3.13 110.7 0.93 0.31 0.27 
4.16 177.3 2.15 0.40 –10.30 
4.39 302.5 3.61 0.60 –47.66 
4.87 381.3 5.48 0.84 –49.22 
4.93 289.0 23.58 2.53 –167.73 
Simulation Description Year Real GDP (%) Employment ('000s) GDP deflator Real wages Current balance (£m) 
£100m p.a. for 1929–31, constant short-term interest rates 3.40 120.3 1.01 0.34 0.29 
4.46 190.5 2.34 0.44 –11.30 
4.87 332.7 3.95 0.65 –47.51 
1.53 270.8 5.40 0.72 –50.62 
0.26 –7.4 7.36 0.74 –52.40 
£100m p.a. for 1929–38, constant short-term interest rates 3.40 120.3 1.01 0.34 0.29 
4.46 190.5 2.34 0.44 –11.30 
4.87 332.7 3.95 0.65 –47.51 
4.98 403.4 5.95 0.90 –50.54 
5.68 321.6 24.45 2.65 –170.03 
£100m p.a. for 1929–31 with crowding-out effects 3.13 110.7 0.93 0.31 0.27 
4.16 177.3 2.15 0.40 –10.30 
4.39 302.5 3.61 0.60 –47.66 
1.42 249.7 4.94 0.66 –49.30 
0.43 9.2 7.64 0.78 –54.53 
£100m p.a. for 1929–38 with crowding-out effects 3.13 110.7 0.93 0.31 0.27 
4.16 177.3 2.15 0.40 –10.30 
4.39 302.5 3.61 0.60 –47.66 
4.87 381.3 5.48 0.84 –49.22 
4.93 289.0 23.58 2.53 –167.73 

Note: Variables as % of base run values, except employment (’000s) and current balance (current prices), which are changes from base values.

Source: Derived from Dimsdale and Horsewood (1995, Table 4).

A deeper analysis of the ‘Treasury view’ on the futility of deficit-financed public works, with the special role played by psychological crowding-out (Middleton, 1985, ch. 8), would counsel even graver caution on the likely viability of the Keynesian solution. As we have seen, such were the debt dynamics that there was a constant risk that markets might require a risk premium for new borrowing if they sensed major deviations from the path of accepted fiscal prudence in peacetime. As it was, with the sinking fund basically suspended from 1931/2 and the authorities placing so much emphasis on cheap money as the centrepiece of the recovery strategy, it was entirely logical to perceive a potential trade-off between an active monetary policy and untried Keynesian fiscal policies. Had orthodox financial opinion been convinced of the effectiveness of the Keynesian solution, financial and business opinion might have been different, but there was no such acceptance at this time nor any in immediate prospect. Moreover, as Wren-Lewis (2010, p. 78) shows for our current troubles, the point at which markets become spooked cannot be established a priori, and: ‘The key limit is not when a government will probably default, but when the possibility that it will default becomes significant enough for lenders to require a risk premium to offset this chance.’ For Britain, in 1931–3 the risk of risk premia would not only have imperilled cheap money but, via the debt dynamics, budgetary equilibrium and thus the delicate balancing act that was the 1931/2 second budget and the 1932/3 ‘gamble’ budget. Thus, however big the fiscal policy multiplier, given the size of the unemployment problem even before the 1929 demand shock, the fact that budget deficits would need to be sustained for a number of years inevitably posed the risk of higher interest rates and thus crowding-out.

Conclusions

We are far from being the first generation to seek lessons from the Great Depression, with the pioneers in this respect being those who designed the post-Second World War international economic order and individual national economic management regimes in the hope of avoiding the worst mistakes of the 1930s. In an aptly titled The Economics Lessons of the Nineteen-Thirties, a comparative study of economic policy in the UK, US, France, Germany, and Sweden, Arndt (1944, p. 221) identified ‘the absence of any effective system of international co-ordination of national economic policies’ as one of the basic problems. Certainly, once the World Economic Conference failed in 1933, each country was largely dependent upon its own devices for securing national recovery. Here, then, is a very basic difference between now and then which conditions how we view the 1930s.

On Britain, Arndt (1944, p. 125) observed: ‘The most striking fact about . . . internal economic policy . . . is that Great Britain was the only one of the five countries . . . which did not resort to a policy of budget deficits to promote internal recovery.’ The passage of time and the accumulation of scholarship, in which the release of the official papers has been key, suggest that Britain’s distinctiveness was somewhat different. Admittedly, it did not pursue Keynesian activism, as in a deliberate strategy of deficit-finance, nor what we might term Keynesian passivity, as in the example of the Hoover and then Roosevelt administrations which ran budget deficits as the automatic stabilizers were not overridden, but the authorities did pursue a very active, and at times rather subtle, recovery policy by the standards of orthodox finance.

The overall effectiveness of macroeconomic policy is difficult to evaluate since current policy objectives and targets are at variance with those of the 1930s. That said, within the limited role accorded fiscal policy as a stabilization instrument, policy can be judged to have been successful because, with the exception of the 1931 crisis, fiscal operations provided a stable environment for business and permitted an expansionary monetary policy—conditions conducive to recovery which, as we have seen, was very much more significant than that in the US. It is when fiscal policy is appraised with reference to modern objectives—those, of course, formulated and developed during the 1930s in response to the limitations and unsatisfactory macroeconomic impact of the existing policy—that its deficiencies and destabilizing characteristics become evident. It is a staple of the literature ‘that recovery proceeded less strongly and rapidly than it would have done with more enlightened budgetary policies’ (Winch, 1972, p. 218). This seems a general lesson of the 1930s, one recently confirmed by Almunia et al.’s (2010) cross-country study of 27 economies which identiifed small monetary policy effects but very much more substantial potential for fiscal policy (short-run multipliers of 2.5; longer-run 1.2), potential not realized in most of their countries: ‘Where significant fiscal stimulus was provided, output and employment responded accordingly. Where monetary policy was loosened, recovery occurred sooner’ (p. 250). But, on the whole, fiscal policy was not used; it was not that it was ineffective.

Romer (2009) has identified six lessons from the US experience of the Great Depression: first, small fiscal expansion has small effects; second, monetary expansion is helpful even when interest rates are near zero; third, beware of cutting back the stimulus too soon; fourth, financial recovery and real recovery go together; fifth, worldwide expansionary policy shares the burdens and benefits of recovery; and, sixth, the Great Depression did eventually end (and would have ended earlier had there not been the 1937 policy-induced setback). The British experience does not wholly map to these lessons as there was no banking crisis and no policy-induced setback (numbers 4 and 6). Additionally, the British case illustrates some more enduring lessons, not least that—as Keynes long argued—it is vital that a recession is never allowed to become a depression. Fortunately, in 2008–9 western policy-makers appeared to understand this lesson, though not all appear to do so now in all countries, not least Britain. The critical role of the exchange-rate regime is also confirmed by the British case, with the securing of a floating exchange rate widening the policy space at a critical juncture, upon which recovery hinged. The lesson here is obvious, though highly contextual; it also leads us back to Bernanke’s fourth lesson, ‘history is never a perfect guide’.

Lessons of a different sort are revealed by the British experience of overriding the automatic stabilizers and of fiscal consolidation. On the former, Keynes had argued at the time, and it has since been confirmed by recent econometric work (Dimsdale and Horsewood, 1995), that crowding-out was not a clear and present danger; instead, the authorities gambled on crowding-in when they overrode the automatic stabilizers. Given the high macro-marginal budget rate, this was a real gamble, and had it gone wrong the likely consequences would have been a vicious circle of an ever tighter fiscal stance which in depressing GDP made impossible ex post budget equilibrium.

The gamble was that fiscal consolidation in 1931 would buy sufficient policy credibility to assuage orthodox opinion and boost business confidence, and it is important that it was motivated in part by the objective of defending the exchange rate. With gold’s abandonment, a successful consolidation was even more important:13

It is one thing to go off the gold standard with an unbalanced budget and uncontrolled inflation; it is quite another thing to take this measure, not because of internal financial difficulties, but because of excessive withdrawals of borrowed capital.

Had the 1931/2 financial year not closed with a budget surplus as then measured (Bc) and 1932/3 had a deficit which could be presented as manageable given that recovery was now under way, the strategy could have failed, and spectacularly so. Two further aspects of fiscal consolidation need highlighting: the first, the extent to which the 1931 crisis was exploited by those seeking to contain Leviathan; and the second that, whatever the rhetoric, retrenchment did not translate into public-sector job cuts. Nominal wage and salary cuts of 10–20 per cent were planned for the public sector, but not wholly delivered after a naval mutiny (Mowat, 1955, pp. 403–6); however, public-sector employment carried on growing, absolutely and relatively, although admittedly from a base very much smaller than today (in 1929, including the armed forces, public employment was 5 per cent of total employment, reaching 5.8 per cent by 1938: Feinstein, 1972, Table 59). As fiscal consolidation, the September 1931 package marked but a temporary halt to public expenditure growth (the Treasury knew this full well, being deeply concerned about the onward march of welfare spending), but, as we have seen, by modern standards the interwar authorities’ conduct was exemplary: set the challenge of debt stabilization in the context of a cyclically unstable economy and a downward price trend, they attained a primary balance which never fell below + 7 per cent of GDP between 1929 and 1932, this at a point when the debt-to-GDP ratio averaged 165 per cent, whereas in 2009/10, at which point on the Maastricht definition the debt-to-GDP ratio was 68.2 per cent, the primary balance was –4.7 per cent of GDP (Barrell and Kirby, 2010, Table 3) and would have deteriorated sharply had not first Labour and then the coalition government accelerated the consolidation path.

In contemporary Britain, the commentariat and policy-makers (informed economists less so) have been prone to seek lessons from the 1930s for the present. In part, this paper has been motivated to show that such lessons are sparse and need careful calibration against the very different situation of 80 years ago. The economic lessons are either basically obvious or so situation-contingent as not to constitute useable knowledge for managing the Great Recession, whereas the political economy lessons are stronger and more generic: to secure big change requires a gamble with the real economy if fiscal consolidation becomes politically more appealing than the economic logic of stabilization.

References

Almunia
M.
, et al.  . 
From Great Depression to Great Credit Crisis: Similarities, Differences and Lessons
Economic Policy
 , 
2010
, vol. 
25
 
2
(pg. 
219
-
65
)
Arndt
H. W.
The Economic Lessons of the Nineteen-Thirties
 , 
1944
Oxford
Oxford University Press
Bank of England
The UK Economic Recovery in the 1930s
 , 
1984
London
Bank of England
Barrell
R.
Kirby
S.
UK Fiscal Prospects
National Institute Economic Review
 , 
2010
July
, vol. 
213
 (pg. 
F66
-
F70
)
Bernanke
B.
Essays on the Great Depression
 , 
2000
Princeton, NJ
Princeton University Press
Bernanke
B.
Economic Policy: Lessons from History
Economic Policy: Lessons from History
 , 
2010
Washington, DC
Center for the Study of the Presidency and Congress
 
8 April, available at www.federalreserve.gov/newsevents/speech/bernanke20100408a.htm
Bernstein
M. A.
The Great Depression: Delayed Recovery and Economic Change in America, 1929–1939
 , 
1987
Cambridge
Cambridge University Press
Bordo
M. D.
Schwartz
A. J.
Taylor
J. B.
Woodford
M.
Monetary Regimes and Economic Performance: The Historical Record
Handbook of Macroeconomics
 , 
1999
, vol. 
1A
 
Amsterdam
Elsevier
(pg. 
149
-
234
)
Broadberry
S. N.
Fiscal Policy in Britain during the 1930s
Economic History Review
 , 
1984
, vol. 
37
 
1
(pg. 
95
-
102
2nd ser.
Broadberry
S. N.
The British Economy Between the Wars: A Macroeconomic Survey
 , 
1986
Oxford
Basil Blackwell
Brown
E. C.
Fiscal Policy in the ’Thirties: A Reappraisal
American Economic Review
 , 
1956
, vol. 
46
 
5
(pg. 
857
-
79
)
Buchanan
J. M.
Wagner
R. E.
Democracy in Deficit: The Political Legacy of Lord Keynes
 , 
1977
New York
Academic Press
Capie
F. H.
Collins
M.
The Inter-war British Economy: A Statistical Abstract
 , 
1983
Manchester
Manchester University Press
Capie
F. H.
Collins
M.
Wood
G. E.
Floud
R. C.
McCloskey
D. N.
Money in the Economy, 1870–1939
The Economic History of Britain since 1700
 , 
1994
, vol. 
II
 
2nd edn
Cambridge
Cambridge University Press
(pg. 
217
-
46
)
Carter
S. B.
, et al.  . 
Historical Statistics of the United States: Earliest Times to the Present, 5 vols
 , 
2006
New York
Cambridge University Press
Chadha
J. S.
Dimsdale
N. H.
A Long View of Real Rates
Oxford Review of Economic Policy
 , 
1999
, vol. 
15
 
2
(pg. 
17
-
45
)
Dimsdale
N. H.
British Monetary Policy and the Exchange Rate, 1920–1938
Oxford Economic Papers
 , 
1981
, vol. 
33
 
2, Supplement
(pg. 
306
-
49
)
Dimsdale
N. H.
Employment and Real Wages in the Inter-war Period
National Institute Economic Review
 , 
1984
, vol. 
110
 (pg. 
94
-
103
(November)
Dimsdale
N. H.
Horsewood
N.
Fiscal Policy and Employment in Interwar Britain: Some Evidence from a New Model
Oxford Economic Papers
 , 
1995
, vol. 
47
 
3
(pg. 
369
-
96
)
Eichengreen
B. J.
Golden Fetters: The Gold Standard and the Great Depression, 1919–1939
 , 
1992
New York
Oxford University Press
Eichengreen
B. J.
Sachs
J.
Exchange Rates and Economic Recovery in the 1930s
Journal of Economic History
 , 
1985
, vol. 
45
 
4
(pg. 
925
-
46
)
Feinstein
C. H.
National Income, Expenditure and Output of the United Kingdom, 1855–1965
 , 
1972
Cambridge
Cambridge University Press
Fishback
P.
US Fiscal and Monetary Policy
Oxford Review of Economic Policy
 , 
2010
, vol. 
26
 
3
(pg. 
385
-
413
)
Foreman-Peck
J. S.
Oliver
M. J.
The Debt Constraint on British Economic Policy and Performance in the 1920s
Studies in Economic and Social History: Essays in Honour of Derek H
 , 
2002
Aldershot, Ashgate
Aldcroft
(pg. 
101
-
18
)
Hansard
Financial Statement by the Chancellor of the Exchequer
Parliamentary Debates (House of Commons)
 , 
1931
September
10
, vol. 
256
 (pg. 
297
-
313
5th ser
Hatton
T. J.
Glynn
S.
Booth
A. E.
The Outlines of a Keynesian Solution
The Road to Full Employment
 , 
1987
London
Allen & Unwin
(pg. 
82
-
94
)
Hatton
T. J.
Thomas
M.
Labour Markets in the Interwar Period and Economic Recovery in the UK and the USA
Oxford Review of Economic Policy
 , 
2010
, vol. 
26
 
3
(pg. 
463
-
85
)
HMSO
Memorandum on the Measures Proposed by His Majesty’s Government to Secure Reductions in National Expenditure
 , 
1931
 
BPP 1930–1 (3952), xviii, 371
Holland
D.
, et al.  . 
Decomposition of Fiscal Deterioration in the OECD
National Institute Economic Review
 , 
2010
, vol. 
213
 (pg. 
F13
-
F18
(July)
Howson
S. K.
Domestic Monetary Management in Britain, 1919–38
 , 
1975
Cambridge
Cambridge University Press
Howson
S. K.
Sterling’s Managed Float: The Operation of the Exchange Equalisation Account, 1932–39
 , 
1980
NJ, Princeton University Press
Princeton
Howson
S. K.
Winch
D.
The Economic Advisory Council, 1930–1939: A Study in Economic Advice during Depression and Recovery
 , 
1977
Cambridge
Cambridge University Press
Kavanagh
D.
Cowley
P.
The British General Election of 2010
 , 
2010
London
Palgrave
Keynes
J. M.
Some Consequences of the Economy Report
New Statesman and Nation
 , 
1931
(pg. 
141
-
145
15 August, reprinted in Keynes (1972)
Keynes
J. M.
The Means to Prosperity
 , 
1933
(pg. 
335
-
66
reprinted in Keynes (1972)
Keynes
J. M.
Collected Writings of John Maynard Keynes, vol IX: Essays in Persuasion
 , 
1972
London
Macmillan
Keynes
J. M.
Henderson
H. D.
Can Lloyd George Do It?—The Pledge Examined
 , 
1929
(pg. 
86
-
125
reprinted in Keynes (1972)
Maddison
A.
Dynamic Forces in Capitalist Development: A Long-run Comparative View
 , 
1991
Oxford
Oxford University Press
Middleton
R.
The Constant Employment Budget Balance and British Budgetary Policy, 1929–39
Economic History Review
 , 
1981
, vol. 
34
 
2
(pg. 
266
-
86
)
2nd ser
Middleton
R.
The Treasury in the 1930s: Political and Administrative Constraints to Acceptance of the “New” Economics
Oxford Economic Papers
 , 
1982
, vol. 
34
 
1
(pg. 
48
-
77
)
Middleton
R.
The Treasury and Public Investment: A Perspective on Interwar Economic Management
Public Administration
 , 
1983
, vol. 
61
 
4
(pg. 
351
-
70
)
Middleton
R.
The Measurement of Fiscal Influence in Britain in the 1930s
Economic History Review
 , 
1984
, vol. 
37
 
1
(pg. 
103
-
6
)
2nd ser
Middleton
R.
Towards the Managed Economy: Keynes, the Treasury and the Fiscal Policy Debate of the 1930s
 , 
1985
London
Methuen
Middleton
R.
Government versus the Market: The Growth of the Public Sector, Economic Management and British Economic Performance, c.1890–1979
 , 
1996
Cheltenham and Brookfield, VT
Edward Elgar
Middleton
R.
Floud
R. C.
Johnson
P. A.
Government and the Economy, 1860–1939
The Cambridge Economic History of Modern Britain
 , 
2004
, vol. 
II
 
Cambridge
Cambridge University Press
(pg. 
456
-
89
)
Mitchell
B. R.
British Historical Statistics
 , 
1988
Cambridge
Cambridge University Press
Mitchell
J.
Solomou
S. N.
Weale
M.
Monthly and Quarterly GDP Estimate for Interwar Britain
National Institute of Economic and Social Research Discussion
 , 
2009
 
Paper No. 348
Moggridge
D. E.
British Monetary Policy, 1924–1931: The Norman Conquest of $4.86
 , 
1972
Cambridge
Cambridge University Press
Mowat
C. L.
Britain between the Wars, 1918–1940
 , 
1955
London
Methuen
Nevin
E. T.
The Mechanism of Cheap Money: A Study of British Monetary Policy, 1931–1939
 , 
1955
Cardiff
University of Wales Press
Obstfeld
M.
Shambaugh
J. C.
Taylor
A. M.
The Trilemma in History: Tradeoffs among Exchange Rates, Monetary Policies and Capital Mobility
Review of Economics and Statistics
 , 
2005
, vol. 
87
 
3
(pg. 
422
-
38
)
OECD
Fiscal Consolidation: Lessons from Past Experience
Economic Outlook
 , 
2007
, vol. 
81
 (pg. 
29
-
52
)
ONS
Government Deficit and Debt under the Maastricht Treaty
Statistical Bulletin, Office for National Statistics
 , 
2010
 
31 March
Peppers
L. C.
Full-employment Surplus Analysis and Structural Change: The 1930s
Explorations in Economic History
 , 
1972
, vol. 
10
 
2
(pg. 
197
-
210
)
Phelps Brown
E. H.
Shackle
G. L. S.
British Economic Fluctuations, 1924–38
Oxford Economic Papers
 , 
1939
, vol. 
2
 
1
(pg. 
98
-
134
)
Pigou
A. C.
Aspects of British Economic History, 1918–1925
 , 
1947
London
Macmillan
Redmond
J.
An Indicator of the Effective Exchange Rate of the Pound in the Nineteen-Thirties
Economic History Review
 , 
1980
, vol. 
33
 
1
(pg. 
83
-
91
)
2nd ser
Renaghan
T. M.
A New Look at Fiscal Policy in the 1930s
Research in Economic History
 , 
1988
, vol. 
11
 (pg. 
171
-
83
)
Richardson
H. W.
Economic Recovery in Britain, 1929–39
 , 
1967
London
Weidenfeld & Nicolson
Romer
C. D.
The Great Crash and the Onset of the Great Depression
Quarterly Journal of Economics
 , 
1990
, vol. 
105
 
3
(pg. 
597
-
624
)
Romer
C. D.
What Ended the Great Depression?
Journal of Economic History
 , 
1992
, vol. 
52
 
4
(pg. 
757
-
84
)
Romer
C. D.
Lessons from the Great Depression for Economic Recovery in 2009
2009
Brookings Institution
 
9 March, available at http://www.brookings.edu/~/media/files/ events/ 2009/0309_lessons/0309_lessons_romer.pdf
Sayers
R. S.
The Bank of England, 1891–1944, 3 vols
 , 
1976
Cambridge
Cambridge University Press
Sefton
J.
Weale
M. R.
Reconciliation of National Income and Expenditure: Balanced Estimates of National Income for the United Kingdom, 1920–1990
 , 
1995
Cambridge
Cambridge University Press
Solomou
S. N.
Themes in Macroeconomic History: The UK Economy, 1919–1939
 , 
1996
Cambridge
Cambridge University Press
Thomas
M.
Rearmament and Economic Recovery in the late 1930s
Economic History Review
 , 
1983
, vol. 
36
 
4
(pg. 
552
-
79
)
2nd ser
Turner
P. M.
Wealth Effects and Fiscal Policy in the 1930s
Economic History Review
 , 
1991
, vol. 
44
 
3
(pg. 
515
-
22
)
2nd ser
Williamson
P. A.
National Crisis and National Government: British Politics, the Economy and Empire, 1926–1932
 , 
1992
Cambridge
Cambridge University Press
Winch
D. N.
Economic and Policy: A Historical Study
 , 
1972
revised edn
London
Fontana
Worswick
G. D. N.
The Sources of Recovery in the UK in the 1930s
National Institute Economic Review
 , 
1984
, vol. 
110
 (pg. 
85
-
93
(November)
Wren-Lewis
S.
Macroeconomic Policy in Light of the Credit Crunch: The Return of Counter-cyclical Fiscal Policy?
Oxford Review of Economic Policy
 , 
2010
, vol. 
26
 
1
(pg. 
71
-
86
)
1
See ‘The Pain to Come’, Economist, 25 March 2010 for a reasoned, comparative assessment; and Kavanagh and Cowley (2010) on economic factors in the 2010 general election campaign.
2
His major essays, solely and jointly authored, are collected together in Bernanke (2000).
3
Mitchell et al. (2009, Tables 2A, 1A). Turning points respectively 1930Q1 and 1932Q3; and January 1930 and—twin troughs—September 1931/July 1932; then 1938Q1 and January 1938.
4
Capie and Collins (1983, Tables 4.4); on a seasonally adjusted basis the national insurance series peaks at 22.5 per cent in August–September 1932 (Table 4.5).
5
Calculations: UK, from Sefton and Weale (1995, Table A.3); US, series Ca84, Ca 87, and Ca88 from Carter et al. (2006, III).
7
See Middleton (1981, pp. 275–8). We lack similar research for the US as existing studies (Brown, 1956; Peppers, 1972; Renaghan, 1988) have not investigated the automatic stabilizing properties of that fiscal system as part of their full-employment budget calculations. Nonetheless, we can presume that the federal deficit was less responsive to income changes because of a lower welfare effort (thus lower unemployment benefits) and less progressive tax structure. This, together with the given that the federal budget was a much lower proportion of GDP than the UK central government budget, suggests much reduced automatic stabilizing properties relative to the UK.
8
This was prompted by Bank of England (1984), an exercise involving a panel of academic consultants who, as the British economy was recovering from the recession of 1979–81, examined the course of the 1930s from the perspective of lessons that might be learnt. Worswick (1984) is a reworking of his panel paper.
9
There is a broad consensus among economic historians about interwar monetary policy, with little new work since the foundational studies of Moggridge (1972), Howson (1975), Sayers (1976), and Dimsdale (1981), though Nevin (1955) is still relevant, and especially for the impact of cheap money on the housing market and the cost of finance for industrial and commercial companies.
10
There has been debate about the appropriateness of this measure of fiscal policy; see Broadberry (1984); Middleton (1984); and Turner (1991).
11
The Economist (‘A New Axe’, 8 August 1931, p. 255) was initially a rare voice of balanced judgement, noting that the figure of £120m ‘rather seriously overpaints the gloom of the immediate budgetary prospect’ as this was not the central government budget deficit per se as it included £40m of prospective new borrowing on the Unemployment Insurance Fund that was not, on current accounting conventions, part of the government budget. Later, The Economist accepted the inevitability of very significant fiscal consolidation which, at the time of the emergency September budget, encompassed a prospective ex ante deficit of £75m for 1931/2 and £170m for 1932/3 (‘Balancing the Budget’, 12 September 1931, p. 460). Additionally, within the Treasury some senior officials considered the May report’s criteria for a balanced budget as excessively severe (Middleton, 1985, p. 113).
12
There is no one publication which combined all aspects of the fiscal consolidation, but see Hansard (1931), HMSO (1931), and Middleton (1985, Table 6.2).
13
Treasury press release, 20 September 1931, cited in Williamson (1992, p. 422).

Author notes

Comments from Nicholas Dimsdale, other participants at the British Academy conference, ‘Lessons from the 1930s Great Depression for the Making of Economic Policy’, 15–16 April 2010, and the referees are acknowledged gratefully, with responsibility for any remaining errors my own.