Useful Pages

Thursday, June 13, 2013

The Profit Deflation of the 1890s

The phenomenon of “profit deflation” in the 1890s is described in this fascinating analysis by H. Clark Johnson:
“The international deflation of 1891–96 directly compressed profits. The extent of actual price decline was less than for the two income deflations considered above. From 1890 through 1896, Sauerbeck’s British wholesale price index declined by 18 percent and The Economist’s index dropped by 14 percent. British money wages, however, actually rose by several percentage points, so the rise in real wages was striking. The rate of investment dropped sharply; new capital issues averaged £102 million during 1880–89 and £154 million during 1889-90 but fell to an average level of £70 million during 1891–96. (These data depict a trend; investment need not be financed through new issues.) The rate of saving was high and increased from perhaps £150 million annually in 1880 to £200 million annually in 1896. Aggregate savings deposits grew greatly during the 1890s, both at the Post Office and at private banks. As investment declined despite the increase in savings, the second term of the price equation turned negative, while the first term increased slightly but steadily — reflecting the rigidity of input costs.

The pattern in the United States was similar. During 1893–96, the wholesale price index declined by 2.4 percent annually, compared to a decline of 1.1 percent annually during 1879–92. Unlike wages during the deflation of the 1870s, hourly wages were steady in nominal terms and hence rose in real terms. (Evidence on British and American wage levels during the 1890s undermines frequent assertions that wages were flexible during the period of the prewar gold standard.) Whereas the (nominal) volume of New York City bank clearings was steady during the deflation of the 1870s, it decreased abruptly during 1892–94. Tobin’s q declined moderately from 1892 through 1896, which was significant in part because it followed a full decade of stagnation in real stock prices. The annualized stock index level of 1881 was not exceeded until 1899.

The 1890s saw intense agitation for inflationary policies, and a central plank of William Jennings Bryan’s Democratic party platform of 1896 was that the gold standard should be abandoned in favor of bimetallism. When the Republicans won the election, the gold standard was again perceived as being secure. This conclusion was soon reinforced by rising world gold output and the beginning of a mild international inflation, which weakened the political attraction of bimetallism.” (Johnson 1997: 20).
There are two issues here, although the second is more important for my purposes:
(1) the idea that the 19th century was a period of relatively flexible wages, and

(2) the effects of the price deflation from 1873 to 1896, and in particular on profits and the level of investment.
First, it appears wages were not as flexible in the 1890s, during this later era of the gold standard, as some economists think.

Secondly, it appears that profit deflation, from the price deflation, with relative wage rigidity, induced a fall in investment. That was part of the economic crisis in the 1890s.

Now some neoclassical Marshallian economists at Cambridge University had their own pre-Keynesian theory about the causes of the late 19th century economic problems in the 1880s.

John Neville Keynes, John Maynard Keynes’s father, gave his own evidence to the UK “Royal Commission on the Depression of Trade and Industry” (whose final report was published in 1886).

He saw price deflation as having the following undesirable effects, as described by Skidelsky:
“These linkages were brought out by Neville Keynes in his evidence to the Royal Commission on the Depression of Trade and Industry (1886). The depression in trade was ‘partly but not wholly due’ to the rise in the value of gold relative to other commodities. This discouraged enterprise for five reasons:
(a) because a fall in price between the start and the completion of a transaction involved the trader in loss;

(b) because the trader tended to exaggerate his own loss by not taking sufficient account of the general fall in prices,

(c) because the profits of enterprise were temporarily diminished on account of increased depreciation of fixed capital;

(d) because the ratio of profits to wages fell as a result of the fall in money wages lagging behind the fall in prices, and

(e) because the fall in prices increased the burden of debt, transferring wealth from borrowers to lenders.
Such evidence was not intended to challenge the now orthodox quantity theory, merely to point to the difficulties of adjusting from one price level to another. Its implication was that monetary policy should be used to raise prices, and thereafter stabilise the price level. Out of such considerations developed the movement for bimetallism, which was an attempt to increase the amount of legal tender money by obliging the central bank to mint both gold and silver on demand at a fixed ratio.” (Skidelsky 1983: 231).
So John Neville Keynes anticipated modern concerns about debt deflation and also identified profit deflation as one of the causes of decreased private investment during this period of deflation.


BIBLIOGRAPHY
Johnson, H. Clark. 1997. Gold, France, and the Great Depression, 1919–1932. Yale University Press, New Haven and London.

Skidelsky, R. J. A. 1983. John Maynard Keynes: Hopes Betrayed 1883–1920 (vol. 1). Macmillan, London.

18 comments:

  1. "(a) because a fall in price between the start and the completion of a transaction involved the trader in loss;"

    I don't see how this could be he case if the fall in price was due to productivity improvements. In this case the supplier chooses to lower prices as a result of lower costs. Other (less productive) suppliers may experience what you describe but their losses would be offset by others gains so this could not be a general phenomenon.

    What is being described is more consistent with a fall in AD related to changes in money demand (bad deflation in Selgin's model)

    ReplyDelete
    Replies
    1. "I don't see how this could be he case if the fall in price was due to productivity improvements. In this case the supplier chooses to lower prices as a result of lower costs."

      So "good deflation" is being defined as any price reduction from productivity growth that does not cause reduced income or profit?

      Yet presumably many types of productivity growth will result in more unemployment and worker income, which is likely to induce a fall in AD.

      Delete
    2. Rob,

      Changes in prices (and costs) do not affect the society homogeneously, so it may be difficult to separate the 'good' and 'bad' deflation. It is never the case that all humans suddenly become more productive at the same type. For example, what for a factory might be a boon (lower costs of labor leading to the lower price of product allowing to expand market share) for the trader might be ruinous if he expected to sell at high price.

      If, for example, the owner of factory lays off half of his workers due to their improved productivity, then it is certain that while searching for a job those workers will have greatly reduced effective demand (sans government transfers) for some indeterminate time. Even if factory owner is better off and workers are able to find work quickly, that simply must affect macroeconomical situation to some extent.

      I don't even think that this is an anti-Austrian idea - especially since Austrian economics always considered heterogeneity of agents and capital and distortions in price systems to be very important.

      Delete
    3. So your envisaging something like a productivity increase for a good whose demand is relatively inelastic leading to less workers being employed in that industry - leading to some sort of AD shock ?

      I guess that's possible (though empirical evidence does not show much correlation between high productivity growth and high unemployment). But remember even if that was the case then an appropriate monetary regime would always allow AD and employment to remain at the optimum level.

      Delete
    4. Rob,

      Assuming you replied to me - yes, I envisage something like this. But my point is not that productivity increases are potentially dangerous(they could be), my point is that things that happen on the micro-level could not be independent of what happens on the larger scale, so 'good' and 'bad' effects of deflation (or, conversely, inflation) are, in effect, inseparable.

      Productivity growth leading to lower prices will make some agents better off - but also some agent worth off, as shown by JN Keynes. No monetary regime could alter this situation quickly enough for AD or employment to remain unaffected.

      Delete
    5. "Productivity growth leading to lower prices will make some agents better off - but also some agent worth off,"

      I agree with this, though for society as a while surely being able to produce more output with the same inputs has got to be good, right ?

      Monetary theory does indeed claim that if the money supply can be adjusted appropriately then the effects of AD shocks (such as you envisage following an improvement on the supply side) can be minimized and full employment maintained

      Delete
    6. Rob,

      There is no such thing as society (c). Well, maybe there is, but I think that evaluating some outcome as 'good' for an abstract entity puts us on a treacherous ground. If most are better off when some are worse off as a result of productivity increase, then how you balance their interests is a question of ethics (and politics). It's probably not a problem to produce a lot more stuff even with higher levels of unemployment, or with most of the population living in increasingly miserable living conditions, but it is usually recognized by the social consensus that some basic rights of population must be prioritized over just producing more.

      The question is, why adhere to such monetary theory? Though, again, it all depends on definitions: since recessions happen basically because agents don't have enough money on their hands, giving them more will in some sense solve their main problem. But that's not what orthodox economical theory usually means by 'increasing money supply'. It's either doing some open market operations to increase bank reserves (for those who believe in money multiplier) or setting an interbank interest rate (endogenous money). But even creating the best conditions for the banking system is not very helpful, as we could see now: even with a lot of excess reserves and interest rates basically at zero AD is still depressed in a lot of countries. I think it is pretty obvious that even setting some optimal monetary regime is not sufficient for quickly correcting market discoordinations. If somebody is unemployed and in debt, what good does him do a 0% interbank interest rate?

      Delete
  2. Rob Rawlings.

    Agreed.

    A decline in investment spending is by definition a decline in MV because NGDP (which is a rough proxy) consists of C+I+G+NX

    Any decline in nominal INCOMES, which include worker and corporate, means demand induced or bad deflation

    ReplyDelete
  3. Its entirely possible for nominal incomes, both corporate and worker, to remain steady or even increase, with a general fall in prices, with businesses making it up on increased volume and productivity.

    Of course this didn't happen during the 1890's because of the gold standard, a suboptimal monetary system if there ever was one.

    NGDPLT (Nominal Gross Domestic Product Level Targeting) allows for good deflation while eliminating or at least minimizing the bad

    ReplyDelete
  4. "(e) because the fall in prices increased the burden of debt, transferring wealth from borrowers to lenders."

    There in no obvious reason why this would decrease AD, and even if it did an appropriate monetary regime would adjust the money supply to prevent it having adverse effects.

    ReplyDelete
  5. "British money wages, however, actually rose by several percentage points, so the rise in real wages was striking."

    I can't work out from your piece whether you think this is a good thing because it makes workers wealthier so they spend more increasing aggregate demand or a bad thing because it decreases profits and hence investment.

    ReplyDelete
    Replies
    1. Why can't it be both at the same time?

      And one effect -- depressed business investment -- also hurt workers by causing higher unemployment?

      The solution was avoiding deflation.

      Delete
    2. That would be ideal, but you are assuming that higher profits necessarily leads to higher real wages. The data does not support this view (Gavyn Davies piece) which shows higher profits leading to increasingly unequal income distribution because lower income workers have not been able to benefit from rising productivity growth. The reality is that nominal wage levels are not determined in a closed economy.

      Delete
    3. Since I support trade unions and rises in wages in line with productivity growth, with income policy, I do not see a problem.

      I am not sure if you are clear on this, but I advocate non-neoclassical, Post Keynesian economics:

      https://en.wikipedia.org/wiki/Post-Keynesian_economics

      Delete
    4. Thanks for the clarification. What's the post Keynesian view on the Phillips Curve?

      Delete
    5. Post Keynesianism obviously doesn't deny that, empirically, as you approach full employment it is likely inflation will increase.

      But the Phillips curve is rejected in Post Keynesianism as too simplistic, partly because of the widespread existence of fixprice/administered price markets.

      Delete
    6. "Post Keynesianism obviously doesn't deny that, empirically, as you approach full employment it is likely inflation will increase."

      Depends how you define 'full employment'.

      You can bring economic expansion to a halt well ahead of any significant supply side tightening, and as long as you have something for everybody unengaged in that expansion to do then you have 'full employment'.

      The solution to the inflation brought about by the Phillips curve is "don't push it that far then".

      Delete
  6. "new capital issues averaged £102 million during 1880–89 and £154 million during 1889-90 but fell to an average level of £70 million during 1891–96."

    The main driver behind the drop off in capital issues from 1890 was due to the collapse of Barings because of their exposure to Argentina. It had severe knock on effect in the London market for the first part of the 90s. You might want to take a closer look at this effect.

    ReplyDelete