The Economic Journal, i o o {March 1990), 176-187 Printed in Great Britain

FISCAL REFORMS AND STABILISATION: FOUR HYPERINFLATION CASES EXAMINED

Gustavo H. B. Franco*

Most accounts, old and new, of the end of the European hyperinflations of the 1920s attribute a key role to 'fiscal reforms' implemented simultaneously with the stabilisations.^ Surprisingly, however, there has been little effort in detailing the contents of these reforms. Indeed, budgets were balanced very quickly after price stability was achieved, but this fact in itself is not actually sufficient to establish that these reforms effectively took place since inflation affects budget deficits in various respects,^ so that the influence of price stability on deflcits might very well be an important part of the explanation for this sudden budgetary improvement. This paper attempts to assess this possibility, and its basic contention is that the presence of such 'reforms' can be disputed, once evidence is provided on the nature of fiscal measures undertaken along with stabilisations and especially on the so called Oliveira-Tanzi effect, i.e. the fact that tax revenues were negatively affected by inflation.* The relevance of alleged tax 'reforms' for the end of the hyperinflations has to be established by distinguishing how much of the observed improvement in tax collection was due to ' reforms' and how much was a consequence of price stabihty. The paper is actually organised around this simple idea. The next three sections examine the fiscal measures undertaken during the stabilisations of Austria, Hungary, Poland and Germany and the last section summarises the main findings, comments briefly on alternative routes to explain these stabihsations. I Although the evolution of economic policy in Austria and Hungary followed very specific paths, these countries share the origin of their problems - the sharp reductions in their territories and populations determined by the dismemberment of the Habsburg empire - and the fact they become basket cases of stabilisation programmes conceived and managed by the League of Nations, a more stringent and interventionist version of its outgrown, the IMF programmes of our days. Indeed, the empire dismemberment has caused many problems, some of them of a fiscal nature. Tables i and 2, reporting annual fiscal accounts of the new repubhcs, provide some indications in this direction.
* Under the usual caveats I would like to thank Gregor Binkert, L. A. Correa do Lago, Barry Eichengreen, Winston Fritsch, Jeffrey Sachs, Lance Taylor, Susan Vitka and two anonymous referees for numerous suggestions on earlier verions of the paper. The author is grateful to the Brazilian National Council for Research (CNPq) for financial support. ' Most notably the recent views of Sargent (1982) and Holtfrerich (1985). " See for example Blejer and Cheasty (1988) and also Tanzi et al. (1987). ' The classic case is Argentina, and the most usual references are Tanzi (1978) and Oliveira (1967). [ 176 ]

[MARCH

1990]

FISCAL REFORMS AND STABILISATION Table i

177

Austria: Closed Accounts Budgets, 1920-4 (millions of gold crowns)

1923 (ist half):

1923 (2nd half) Estimate 188-7 269-7 80-7 70-0 Actual '923II '924II 272-8 296.7 23-9 91-9 483-3 623-4 593-2 632-4 '09-9 9-0% 8.-5 98-6

1920* 1921* '922t Revenues Expenditures Deficit Rev./Exp. 211-7 281-4 215-I 567-5 673-8 672-5 355-8 392-4 457-4 37-3 41-8 32-0

Estimate** '58-4 299-5 141-1 52-9

Actual 210-5 296-5 86-0 71-0

Sources and observations: * Fiscal years of 1919/20 and 1920/1, July to June, from Van Sickle (1931, pp. 66 and 72). t League of Nations, estimate for October of 1922 annualized, from League of Nations (1926 a, p. 33). : Ibid p. 49. Difference between values for 1923 full year and 1923 first half II Full fiscal year, January-December, ibid. pp. i i o - i . ** Original estimates of League's programme. All values converted into gold crowns with annual average exchange rates from Walre de Bordes {1924, pp. i'4-39).

Table 2
Hungary: Closed Accounts Budgets: ig2O-^ (millions of gold crowns)
1924-5 (ist half)t 1920-1* 1921-2* .922-31 '923-4t Expenditures Revenues Deficit % Rev./Exp. 574-' 192-4 387-5 33-5 386-4 226-2 160-2 58-5 242-8 170-8 72-0 70-4 409-5 192-6 216-9 47-0 Estimate 186-3 '43-8 42-5 77-2 Actual
205-9 208-0 20 IOI-O

1924-5 t: Estimate 393-9 293-8 100-1 74-6 Actual 422-8 453-' -30-3 107-2

Sources and observations: * Fiscal years from July to June, nominal values from Ecker-Racz (1933, p. 79) defiated with annual average exchange rates computed from price level indexes from ibid. pp. 61-2. t Estimates from League's programme reproduced in Pasvolski (1928, p. 322). : Full year.

It is interesting and somewhat surprising to note that in both cases the levels of expenditure before and after the stabilisations - which was achieved in the end of 1922 in Austria and in the beginning of 1924 in Hungary - were roughly similar. Their composition, however, was markedly different in both cases reflecting important adjustments of these economies to their new economic realities. In the Austrian case, food subsidies, the operational deficit of state railways, and administrative expenses associated with the excessive number of government officials were the items weighting more heavily on expenditures. Two years after the League's intervention, during the fiscal year of 1924, the share of administrative expenses of total expenditure was reduced from 40-4 % in 1922 to 29-6%, but dismissals, which up to December of 1924 reached 71,349 officials, implied an increase in the expenditures with pensions, whose share rose from 6-t % of expenditure in 1922 to i3"5% in 1924. The railways'

178

THE ECONOMIC JOURNAL

[MARCH

deficit share was cut by half, reaching only i2-i % of total expenditure in 1924 and most of it (63-2 %) corresponded to investment expenditure in electrification programmes (League of Nations, 1926a, p. 33). In parallel there has been some significant increases in expenditure for social overhead. A similar phenomenon is observed in Hungary, namely the League programme marked significant shifts in the direction of expenditures with little, if at all, impact in the level of expenditures. The League's plan for Hungary prescribed budget cuts in several directions in addition to reforms in the administration, but from the beginning the League officials made it very clear that a net increase in expenditure should be expected as the programme was implemented. Like in Austria, the League insisted on dismissals - some 10,000-15,000 officials in this case - but understood that the salaries of the remaining public officials, which numbered some 160,000, 'had so shrunk in value it appeared neither possible nor compatible with the interests of the State to refuse to take steps to increase them' (League of Nations, 1926 b, p. 98). Savings would be obtained with respect to state undertakings and monopolies, especially with respect to tbe operational deficit of the state railway system. But as observed in Austria, these savings would be mosdy offset by increased expenses with pension payments and investment outlays (p. i n ) . For both countries, very significant increases in revenues explain the rapid balancing the budget along with stabilisations. In Austria state revenues more than doubled in real terms from 1922 to 1923 while the League's programme had predicted an increase in revenues of only 3 3 % for the fiscal year of 1923 coming mostly from 'realistic' pricing policies for state undertakings and monopolies, and only to a lesser degree from new taxation; rates of direct taxation were actually reduced (Van Sickle, 1931, p. 195). Unexpectedly, however, the new taxes failed to be effectively implemented and revenues from monopolies were far below target; the yield of direct taxation and of Table 3 Austria and Hungary: Monthly Budgets (for the first reconstruction period)
Austria f

Hungaryt
Deficit 24-3 30-3 16-1
0/

Month*
I 2

Expend. 45-4
52-0 52-0

Revenue
21-1 21-7

/o 46-5 4'-7 69-0 98-0 82-5 82-5 70-1

Expend. 30-5 33-8 36-7 33-7 34-0 37-2 205-9

Revenue
"7-9 29-5 30-9 35-6 49-2 44-9 2080

Deficit -12-6 -4-3 -5-8

'/o

3 4 5

5"-'
49-2 46-8 296-5

35-9 50-1 40-6 38-6

2O8-O

ro 8-6
8-2

"-9
"5-2

58-7 87-3 84-2 105-6 '44-7

7-7
2-1

Totals

-88-5

I20'7 ioro

Sources and observations: * For Austria month i is January of 1923 and for Hungary is June of 1924. t Collected directly from the Monthly Reports of the Commissioner-General, deflated with monthly average exchange rates from Walre de Bordes (1924, pp. 114-39). J In gold crowns from Young (1925, vol. II, p. 326) and from the provisional budgets collected from the Monthly Reports of the Commissioner Ceneral. The values differ slightly from those of Table i due to different methods of transforming paper crowns into gold crowns.

1990]

FISCAL REFORMS AND STABILISATION

179

commodity taxes, in contrast, more than doubled the League's original estimates (League of Nations, 1926a, p. 50; Van Sickle, 1931, pp. 203-4). Table 3 shows that for the first six months ofthe League's intervention revenues were, on average, about 70 % of expenditures and in the second semester this was raised to numbers between 82-5% and 98-0% of expenditures, which may be indicative of the size of the Austrian fiscal deficit correcting for the effect of inflation on tax yields, or if we control for the Oliveira-Tanzi effect. In the Hungarian plan the League's experts seemed more aware of the importance of the effects of inflation on the real yield of taxation, as they referred to the 'proved results ofthe Austrian experience' and argued that 'a substantial part' ofthe required increase in revenues 'may be expected from the automatically better returns (in terms of gold value) from existing taxes (as in the case of Austria) when the stabilisation of the currenty is effected' (League of Nations, 1926 A, pp. 64-6). Yet, the League's experts could never expect a full repetition of the Austrian experience, and thus they devised some scope for fresh taxation.A new land tax was introduced, but its contribution to total revenues during the first reconstruction period was only 3-7% (p. 112). All other taxes remained unchanged at least during the first six months of the programme. As in Austria, revenues were grossly underestimated, especially for the first six months ofthe programme, so that even with the expenditure targets being exceeded by nearly 10 %, the budget was balanced in September of 1924, the fourth month of the scheme, as seen in Table 3. Again the main contribution to this performance came from taxes that the programme left untouched, or reduced, such as the turnover and commodity tax and customs duties,* and mostly as a result of the Oliveira-Tanzi effect. In sum, the stabilisation programmes determined important shifts in the direction of expenditure in both countries which actually represented a significant part ofthe adjustment of these economies to their new frontiers: the gradual resolution of issues like surplus officials and inefficient railways did not have a significant impact on the levels of expenditure, because dismissals were compensated by pension payments and by badly needed increases in real wages,* because of the new demands for social overhead and also because railways efficency required economies but also extensive investment outlays. On the revenues side, a key role would be played by the Oliveira-Tanzi effect, which was recognised in the League's final reports on both programmes, according to which 'budget equilibrium followed [the stabilisation], it did not precede' (League of Nations, 1926a, pp. 75-6 and 19266, p. 37).

II Poland had lost its independence a century before in the Congress of Vienna, when she was partioned among the Russian, German and Habsburg empires. In 1919 Poland was reunified, but with frontiers very different from the ones
* Several imports prohibitions were substituted by tariffs during 1924, which, in addition to a moderate recovery of imports explains part of the recovery of customs revenues. ' In 1923 middle grade government officials' wages stood at 56% of their 1914 values. See International Labour Office (1925, p. 93).

l8o

THE ECONOMIC JOURNAL

[MARCH

Table 4
Poland: Closed Accounts Budgets, 1921-4 (millions of zloty)
1922 '923

1924 II
i75->

1925 II 831-6 986-4 154-8 84-3 96-9 37-4* 76-2

16-2

I
Revenues Expenditure Deficit % Rev./Exp. Property taxj Invest, expandf Coins/notes Debt 1 1

II 43-9 324-9

I
203-4

I 539-9
620-4 80-5 87-0 92-6

I 818-2 9260 107-8 88-4 34-6 n.a. 185-6 5-9

II
800-2

2477
322-5

84-8 74-5 74-8 58-5

16-3

i8ro 44'3
10-5 71-3 12-4

488-0 284-6 58-3

129-8

473'5 298-4 37-0

2-0

84-3
1-8

44'5 *
76-1

895'4 95'2 89-4 23'9 n.a.

115-3 15-2

33-7

76-3

Sources and observations. * From the preliminary accounts. t For 1922 consists of the Extraordinary Property Tax enacted by Michalski and for the later periods consists of Grabski's Capital Levy. X Refers mostly to state railways. Includes small notes and subsidiary coins. II Includes internal and external debt. From Republic of Poland (1926, pp. 173-6 and 1931, p. 80-3).

Table 5
Poland: Monthly Budgets, December of ig2j to May 1924 (millions of zloty)
Property tax December 1923* January 1924 February March April May
1-6 1-8

All other revenues 30-5 32-8 48-0 70-5 103-2

Total revenues
32-1

Invest expendit.*
42-1

Total expendit.
96-0 70-3

% Exp./ revs. 33-4

50-6

35-6
76-0 107-2 121-2 990 100-9

6-9
O-I

28-0

85-5
109-1 107-3 05-5 142-7

36-7
18-0

4-3
3-8

June

947 97-

3-8 9-5 9'

15-1

88-8 98-3
112-9

93-8
70-7

Source and Observations: Preliminary figures from Republic of Poland (1926, pp. 173-6). The totals reported for investment expenditure are not reported in the later closed accounts from which the totals for expenditure were taken. t Grabski's capital levy. From Republic of Poland (1931, pp. 80-3).

of 1815; the new Poland was by all means an entirely new country. The comparison between the first budgets of the new republic and those following the stabilisation, shown in Table 4, reveals that the Polish levels of expenditure nearly tripled in real terms between 1922 and 1925. This increase expressed the overall growth and development of the Polish public sector, which was being built out of the ruins of the bureaucracy left out by the three partitioning empires. It also reflected the pressures generated by the needs of reconstruction: Poland suffered severe devastations in her territory from the war with Soviet Union, which lasted until 1920. The influence of inflation on state revenues was very clearly observed by

FISCAL REFORMS AND STABILISATION

l8l

contemporaries,* but even so the stabilisation programme that started in December of 1923 decided to anticipate the collection of the new capital levy introduced in June of 1923 and scheduled to be collected during 1924-6. The latter represented an important contribution to state revenues especially during the first months of the programme, as shown in Table 5, but yet again the recovery of the yields of existing taxes played the most important role to the observed budgetary improvement (Rose, 1924, p. 11). Little was accomplished in the expenditure side; general savings and expenditure cuts were proposed as part of the usual rhetoric of austerity and sacrifices, but the observed increase in public expenditure from 1923 to 1924 goes opposite to the discourse. The remarkable fact about the role of Polish finances in the process of stabilisation is that despite the very significant improvement, the budget was not balanced as a consequence of the stabilisation programme. It is true that the government was separated from the central bank, but it is surprisingly neglected that the government retained powers to print small notes and coins' that were extensively exercised in 1924 and 1925. During those years the total issuing of money by the Treasury reached approximately 450 million zloty representing more than twice the existing money supply at the onset of stabilisation. Interestingly, however, the increases in debt and in the issue of money by the Treasury during 1924 and 1925 entered the Polish budgetary accounts as 'extraordinary revenues', an accounting trick which was noted by several historians of the episode*, but confused a number of others such as John Parke Young*, Ragnar Nurske^", and more recently Thomas Sargent". Indeed, the failure to take account of the true fiscal situation of Poland during 1924 and 1925 seriously undermines these authors' views on the Polish stabilisation which are centred on the alleged move to a balanced budget. In August of 1925, after a year and a half of stable prices, the zloty was allowed to float and was then stabilised again at a lower level. The London financial press attributed the devaluation to the 'coin inflation' (Mlynarski, 1926, pp. 3-4), but in general it was widely accepted that the collapse was determined by a number of adverse circumstances including harvest failures, weak terms of trade, a tariff war with Germany, the dismal results of the Dillon and Reed loan and the deterioration in competitiveness determined by the wage inflation observed after the stabilisation. Nurske himself denied the influence offiscalpolicy on the episode (League of Nations, 1946, p. 26) and
For example Young (1924, pp. 23-4). It was also noted by the summit of former ministers of finance, who gathered in mid 1923. See Strasburger (1924) and Zweig (1944, p. 36). ' The Treasury was authorised to issue coins up to 150 million zloty for the fiscal year of 1924, representing nearly half of the existing money supply, and for 1925 this total was doubled, Zdziechowski (925. PP- 45-6)' For example Robin {1932, p. 35). See also Heilperin (i 931, pp. 135-6); Zweig (1944, p. 40) and Landau and Tomaszewski (1984, pp. 278-9). Young (1925, vol. II, pp. 182-4). The source of Young's figures was the Polish Statistical Annuary of 1923, published in 1924, so that it could not bring but budget estimates. ' League of Nations (1946, p. 26). " Sargent (1982, pp. 71-2) reproduces a table with budgetary data from Young (1925, vol. II, p. 65), though without mentioning that the figures for 1924 and 1925 correspond to the budget proposed by the government (not even the budget effectively passed in the case of 1925), thus very different from the closed accounts budgets shown in Table 4.

l82 THE ECONOMIC JOURNAL [MARCH Sargent's explanation for the 1925 episode does not involve these issues: referring to the second part of the League of Nations report on post-war inflations the part not authored by Nurske he argued that the collapse had been due to the 'premature relaxation of exchange controls' (Sargent, 1982, p. 73), though the existence of the latter is not mentioned anywhere in the literature'^. In addition, Sargent mentioned 'the tendency of the central bank to make private loans at insufficient interest rates' (p. 73), but in the League's report there is no mention of that. The report only argued that there had been no credit contraction corresponding to the reserves losses after April of 1925, as prescribed by the so called 'rules of the game'. It says nothing about credit at subsidized rates, and again no mention of that could be found elsewhere.

Ill The 'emergency' fiscal decrees of October n t h and December 7th and 19th - which corresponded to the 'reforms' to which the German stabilisation has been usually attributed - were attempts to place the existing tax system 'on a gold basis' and did not introduce any new taxes (Republic of Germany, 1924, pp. 74-9B; Robert, 1926, pp. 137-44 ^"^ Bresciani-Turroni, 1937, pp. 67-74). Indeed, the astonishing growth in tax revenues after the stabilisation observed in Table 6 was generated by the existing taxes (Bresciani-Turroni, 1937, P- 357)In the German case too we observe that the levels of expenditure before and after the stabilisation were very similar, the improvements to the budgetary situation being almost entirely due to increased tax revenues. Thus the 'reforms' observed in Germany had the same character of the ones that have been implemented by the League in Austria and Hungary, namely they were predominantly changes in the composition of expenditure with little effect on the level of expenditures. In this respect it is remarkable that in 1922/3, for example, payments on account of the Treaty of Versailles represented 38% of total expenditure while for 1924/5 only 13% of total spending was budgeted to this purpose, due to the revision of the London Schedule of reparations payments accomplished by the Dawes Plan (Republic of Germany, pp. 32, 77). The service of the domestic public debt was another item that had its share over total expenditures significantly changed, especially during the two years preceding the period covered by Table 7. During 1919 and 1920, when prices grew by approximately 900 %, the real value of the stock of the public debt fell from 58,515 million gold marks - which represented more than seven times public expenditures in 1919-20 - to 5,424 million (pp. 29-32). Already in the fiscal year of 1921-2 the service of this debt claimed only 7% of total expenditures (p. 32 and Statistisches Reichsamt, 1923, p. 42), and in 1923-4 it was reduced further to 3 % (Republic of Germany, 1924, p. 77).
" The League's report mentioning of' foreign trade controls' appears to relate to what in the modern jargon would be termed commercial policy, which is consistent with all accounts of the 1925 episode. League of Nations (1946, p. 108).

1990]

FISCAL REFORMS AND STABILISATION

183

Table 6
Germany: Revenues and expenditures, May of ig2j to October of 1924

(millions of gold marks)

Period "923 May Expenditures Revenues Rev. %/Exp.

June July August September October

November i-io 10-20 20-30

284-7 496-4 473-9 883-6 689-2 235-8

134-2 28-5 258-7 421-4

"23-3 48-2 48-3

11-7 14-4
1-2 01

43-3 9-7
10-2

"-3
2-1

0-5
O-I

Total
December i-io 10-20 20-31 Total "924

'79-9 "65-7 '53-8 499-4

0-4 10-6 I II 32-8 42-9

"-5 4"
2-6

18-2 25-9 57-B 32-9

88-8
164-5

January
I-IO 10-20 20-31

63-9
180-4

Total February March April May June

'99' 443-4 478-6 485-6

472-1 511-0

440-9

98-4 '53-6 '85-9 437-9 339-8 526-8 396-4 449-5 382-3

154-0

85-. 93-3
99-0 71-0 108-4

83-9 88-0 86-7

Sources: Original paper mark figures on a io days basis from Thelwell (1924) p. 28 and (1925) p. 33, deflated weekly exchange rates against sterling from The Economist, various issues (1923).

Table 7
Germany: Revenues and Expenditures, 1921-5 (millions of gold marks)
Yea Expenditure Revenues Deficit % rev./exp.
1921/2* 6,651-3 2,927-4 1922/3* 3.950-6 1,488-1 2,462-5 1923/4*1 '924/5

5,768-0
1,802-5

3.723-9
44-0

3.965-5
31-2

6.895-0 7,786-2 -39"-2

112-9

37-7

Sources and observations: * The values for 1921/2, 1922/3 and for the period March-July of 1923 correspond to the figures reported in Republic of Germany (1924, p. 32). t Thefiguresfor August-December of 1923 are from Table 6. For January-March of 1924 thefiguresare also from Thelwell (1924, p. 33 and 1925, p. 30).

184

THE ECONOMIC JOURNAL

[MARCH

Another important development was the increase in expenditure for broadly defined social purposes - a natural development within the framework of the Weimar state - that actually maintained a strong upward tendency throughout the 1920s (Andic and Veverka, 1964, p. 237). It has often been observed that a decree of October 27th, 1923 established that approximately 25% of all public employees would be dismissed, 10% by January of 1924 (Republic of Germany, 1924, p. 78). It has been less often emphasised that the wages ofthe remaining officials were increased by 50% to 90% in real terms after the stabilisation, very much like what happened to the real wages of other categories of workers,^* so that the net result of these measures on the public service's wage bill might not have been significant. In sum, it seems clear that the deficit was eliminated mostly by the effect of price stability on tax revenues; whatever minimal overall fiscal restraint, real or rhetorical, would serve no purpose other than speeding the strict balancing of the budget, which might have been very useful for the negotiations towards foreign financial support by reassuring the government's adherence to orthodox finance.

IV
The fiscal 'reforms' introduced during the stabilisations comprised mostly shifts in the direction of expenditure with little impact, if at all, on the net budgetary result. The dismantling of the overstaffed bureaucracies of the former empires, the fact that domestic public debts were destroyed by inflation and the external debt in the German case was drastically reduced by the Dawes Plan, did not represent meaningful reductions in public expenditure due to many new influences acting on the contrary direction. The marked increases in government expenditure for social purposes, and the recomposition of wages of public employees, more than made up for the savings often quoted as indications of strict austerity oriented fiscal policies. The recovery of the real yield of taxation after the stabilisation through the Oliveira-Tanzi effect was the key to budget balance in all four cases examined. This means basically that budget deficits were to an overwhelming extent products of inflation, or that, to use today's terminology, 'inflation corrected deficits' were balanced or at least manageable. 'True' tax reforms had taken place at some point before the stabilisations, most likely during previous stabilisation attempts such as Erzberger's in 1921 Germany, Hegedlis in 1921 Hungary and Michalski's in 1922 Poland. All these programmes were specifically designed to address the fiscal issue, but the fact that they failed to arrest inflation does not imply that they failed in their purposes, it suggests instead that fiscal balance was not a sufficient condition for stabilisation at a moment when other fundamental causes of inflation were still in full work. Indeed, one should note that simple causal relations between the hyper" In general the real wages of all categories recovered very quickly from the trough they reached at the end ofthe inflation regaining their pre-war levels late in 1924, Bry (i960, p. 62) and International Labour Office (1925, pp. 16-7). As regards government officials specifically estimates for their levels of real wages, asa percentage of 1914, in October of 1923 are 44% (skilled) a n d 6 i % (unskilled). In June 1924 these levels would be 87% and 99% respectively.

1990] FISCAL REFORMS AND STABILISATION 185 inflations and the fiscal disequilibria would tend to oversimplify the fact that these countries faced huge problems of adjustment to the new economic realities. The redrawing of Central Europe's economic and political map which turned into isolated pieces a customs union that endured for a century - deliberately penalised Austria and Hungary, and created a new country Czechoslovakia - meant to be, for geopolitical reasons, a 'local champion', having received 'a disproportionally large share of the former Monarchy's economic potential' (Berend and Ranki, 1974, p. 182; Bloomfield, 1984, p. 252; and Pasvolski, 1928, p. 329).^* Austrians, for example, would seriously question their 'viability' as a country in the early i92OS^^ when adjustment problems would appear especially dramatic given that international capital markets were closed, the political stability was uncertain and protectionism was generalised in the Danubean region. Germany was a special case in this regard since she was not badly hit by territorial changes, but by the reparations obligation. In fact one can hardly dissociate the latter with the hyperinflation if we consider that Germany was transferring nearly 80 % of her exports - or nearly a quarter of her GDP - as reparations.^* This point is often missed in view of the misleading presentation of the issue made by Machlup (1976), who questions the 'severity of the German transfer problem' (p. 385) on the basis of figures for 1924-8, i.e. after the Dawes Plan had written off a great part of the reparations debt, reduced the annual burden to mere 3-6% of exports in 1924 and to 10-9% on average for 1925-8, and provided a large loan to help recycling these amounts. These transfer problems were made even harder by another circumstance, namely a very strong pressure exercised by workers to recover 1914 real wages. In the beginning of 1920 real wages in these countries stood between half and two thirds of 1914 levels (International Labour Office, 1925, pp. 13-7) and at the same time the labour movement was extraordinarily strengthened by the formidable increase in trade union membership: in Germany the 2,437 thousand enrolled in 1914 turned into 9,163 thousand in 1920. For Austria and Hungary these numbers rose from 147 and 107 thousand in 1914 to 901 and 722 thousand (in 1919) respectively (International Labour Office, 1921, pp. 2-3 and Bry, i960, p. 32). The combination of balance of payments problems and a wage push of very significant proportions (and under flexible exchange rates and currency substitution!) most likely contributed very significantly, if not decisively, to inflation. This is certainly not the place to explore these connections further; our purpose is merely to outline the problems involved and suggest that, regarding 'fundamentals', the necessary conditions for these stabilisations comprised many problems outside the fiscal sphere. Moreover, as
" It is not surprising, given this background, that the Czech could show strong balance of payments position as early as in 1920, and could stabilise their currency in 1921. " For example Pasvolski (1928, p. 108). ^* Keynes (1922, p. 51) reports that for every six months between May and October of 1921 the amounts paid represented 79-6 % of exports revenues of this same period. Applying to the values for exports for 1923 and 1924 the rules established by the 'London Schedule', namely 26% of exports plus 2,000 million gold marks, one obtains ratios of 76-4% and 65-7% respectively. However, these values were not actually paid as Cermany entered into a partial moratorium in the second half of 1922.

l86

THE ECONOMIC JOURNAL

[MARCH

put by Dornbush (1985, p. 12), 'the question of how stabilisation was achieved is not exactly the same as that of why hyperinflation occurred in the first place'. In this respect the process of ' doUarisation' played a crucial role as a coordination device for pricing decisions, as recently observed.^' Indeed, these stabilisations involved the combination of many elements, most likely none of which sufficient in itself.
Rio de Janeiro Date of receipt of final typescript: June ig8g
REFERENCES

Andic, S. and Veverka, J. (1964). 'The growth of government expenditure in Germany since the Unification.' Finazarchiv, vol. 23, no. 2, pp. 198-243. Berend, I. and Ranki, G. (1974). Economic Development in East-Central Europe in XlXth and XXth Century. New York: Columbia University Press. Blejer, M. I. and Cheasty, A. (1988). 'High inflation, heterodox stabilization, and fiscal policy.' World Development, vol. 16, no. 8, pp. 867-81. Bloomfield, J. (1984). 'Surviving in a harsh world: trade and inflation in the Czechoslovak and Austrian Republics, 1918-1926.' In The Experience of Inflation, Intemational and Comparative Studies (ed. G. Feldman et al.). Berlin: Walter de Gruyter. Bresciani-Turroni, C. (1937). The Economics of Infiation, a Study in Currency Depreciation in Post-War Germany. London: Augustus M. Kelley Reprint, 1968. Bry, G. (i960). Wages in Germany, 1871-1945. Princeton: NBER-Princeton University Press. Dornbush, R. (1985). 'Stopping hyperinflation: lessons from the German Experience of the 1920s' NBER Working Paper no. 1675. Ecker-Racz, L. L. (1933). 'Thefinancial reconstruction ofHungary.' Unpublished Ph.D thesis. Cambridge: Harvard University. Franco, G. H. B. (1987). 'The rentenmark miracle.' Revista di Storia Economica, Second Series vol. 4, International Issue, pp. 96-117. Heilperin, M . A . (1931). Le Probleme Mone'taire D'Apre's Guerre et sa Solution en Pologne, en Austriche et en Tchecoslovaquie. Paris: Remeil Siney. Holtfrerich, C. L. (1985). 'Germany and other European countries in the 1920s.' In Infiation and Indexation: Argentina, Brasil and Israel (ed. J. Williamson). Washington: Institute oflnternational Economics. International Labour Ofifice (1921). The Growth of Trade Unionsim During Ten Years, igio-igig. Geneva: International Labour Office. ('925). Workers' Standard of Living in Countries with Depreciated Currency. Geneva: International Labour Office. Keynes, J. M. (1922). A Revision of the Treaty. London: Macmillan-St. Martin's Press, 1971. Landau, Z. and Tomaszewski, J. (1984). 'Poland between inflation and stabilization, 1924-1927.' In The Experience of Inflation, Intemational and Comparative Studies (ed. G. Feldman et al.). Berlin: Walter de Gruyter. and (1985). The Polish Economy in the Twentieth Century. London: Croom Helm. League of Nations (1926(1). Financial Reconstruction of Austria, General Survey and Principal Documents. Geneva: League of Nations. (1926 A). Financial Reconstruction of Hungary, General Survey and Principal Documents. Geneva: League of Nations. (R. Nurske) (1946). The Course and Control of Inflation, a Review of the Monetary Experience of Europe After World War i. Geneva: League of Nations. Lopes, F. L. (1986). 0 Choque Heterodoxo. Rio de Janeiro: Campus. Machlup, F. (1976). 'The transfer problem, theme and four variations.' In Intemational Payments, Debt and Gold. New York: New York University Press. Merkin, G. (1982). 'Towards a theory of the German inflation: some preliminary observations.' In The German Infiation Reconsidered, a Preliminary Balance (eds. G. D. Feldman et at.). Berlin: Walter de Gruyter. Mlynarski, F. (1926). The Intemational Significance of the Depreciation of the Zloty in 1925. Warsaw: The Polish Economist. Oliveira, J. H. G. (1967). 'Money, prices and fiscal lags: a note on the dynamics of inflation' Banca Nazionate del Lavoro Quarterly Review no. 82, pp. 258-67 (September). " See for example Merkin (1982), Lopes (1986) and Franco (1987).

FISCAL REFORMS AND STABILISATION

187

Pasvolski, L. (1928). Economic Nationalism in the Danubean States. New York: MacMillan & Co. Republic of Germany (1924). Germany's Economy, Currency and Finance. Berlin: Kriegslastenkommission. Republic of Poland (1926). Annual Statistics of the Ministry of Finance, 1924. Warsaw: Ministry of Finance. (1926). Annual Statistics of the Ministry of Finance, igz^-igzy. Warsaw, Ministry of Finance. Robert, J. M. (1926). Depredation de la Monnaie et Equilibre Budgetaire: Etude sur les Finances Allemandes 1922-1923. Paris: Les Presses Universitaires de France. Robin, P. (1932). Le Reforme Mone'taire en Pologne. Paris: Marcel Giard. Rose, E. (1924). The Progress of Financial Reconstruction in Poland. Warsaw: Guicodial Publications. Sargent, T. (1982). 'The end of four big inflations.' In Inflation: Causes and Effects (ed. R. Hall) Chicago: NBER-Chicago University Press. Statistisches Reichsamt (1923). Germany's Economy and Financial Situation: An Exhibit of After-Effects of the World War. Berlin. Strasburger, E. (1924). 'A propos du programme de la conferance des anciens ministres du tresor en vue d'assainir les finances polonaises.' VEst Europeen, vol. 4, no. 2, pp. 120-30 (February). Tanzi, V. (1978). 'Inflation, real tax revenue, and the case for inflationary finance: theory with an application to Argentina.' IMF Staff Papers vol. 25, no. 3, pp. 417-51 (September). et al. (1987). 'Inflation and the measurement of flscal deficits.' IMF Staff Papers vol. 34, no. 4, pp. 711-38 (December). Thelwell, J. W. F. (1924). Report on the Economic and Financial Conditions of Germany, revised up to April 1924. United Kingdom: Department of Overseas Trade. (1925). Report on the Economic and Financial Conditions of Germany. United Kingdom: Department of Overseas Trade. Van Sickle, J. V. (1931). Direct Taxation in Austria. Cambridge: Harvard University Press. Walre de Bordes, J. van (1924). The Austrian Crown. London: P. S. King. Young, E. H. (1924). Report on the Financial Conditions of Poland. London: Waterlow & Sons. Young, J. P. (1925). European Currency and Finance. Washington: Commission of Cold and Silver, United States Senate. Zdziechowski, C. (1925). The Finances of Poland, 1924-1925. Warsaw. Zweig, F. (1944). Poland Between the Wars. London: Seeker & Warburg.