What was the scope for British Keynesian policies in the 1930s?

There is a new paper by Nicholas Crafts & Terence Mills in the December 2013 Journal of Economic History, entitled Rearmament to the Rescue? New Estimates of the Impact of “Keynesian” Policies in 1930s’ Britain.  The abstract is here:

We report estimates of the fiscal multiplier for interwar Britain based on quarterly data, time-series econometrics, and “defense news.” We find that the government expenditure multiplier was in the range 0.3 to 0.8, much lower than previous estimates. The scope for a Keynesian solution to recession was less than is generally supposed. We find that rearmament gave a smaller boost to real GDP than previously claimed. Rearmament may, however, have had a larger impact than a temporary public works program of similar magnitude if private investment anticipated the need to add capacity to cope with future defense spending.

You will note that previous estimates of the multiplier for this period are much higher, but this conclusion is based on a new quarterly gdp series for the UK and also on identified “defense shocks,” drawn from The Economist magazine.  At this time, by the way, Great Britain was very close to the zero lower bound and had significant unemployed resources.  I am not sure I would push this as “the correct answer,” but rather as a more general lesson about the fragility of our knowledge in this area.

The published (gated) version of the paper is here.  There are ungated versions here.


'The scope for a Keynesian solution to recession was less than is generally supposed. We find that rearmament gave a smaller boost to real GDP than previously claimed. Rearmament may, however, have had a larger impact than a temporary public works program of similar magnitude if private investment anticipated the need to add capacity to cope with future defense spending.'

There is actually a Republican response to this, especially the entire perspective concerning the value of rearmament -

'Every gun that is made, every warship launched, every rocket fired signifies, in the final sense, a theft from those who hunger and are not fed, those who are cold and are not clothed.

This world in arms is not spending money alone. It is spending the sweat of its laborers, the genius of its scientists, the hopes of its children. The cost of one modern heavy bomber is this: a modern brick school in more than 30 cities. It is two electric power plants, each serving a town of 60,000 population. It is two fine, fully equipped hospitals. It is some fifty miles of concrete pavement. We pay for a single fighter with a half-million bushels of wheat. We pay for a single destroyer with new homes that could have housed more than 8,000 people. . . . This is not a way of life at all, in any true sense. Under the cloud of threatening war, it is humanity hanging from a cross of iron.' http://en.wikipedia.org/wiki/Chance_for_Peace_speech

Of course, that speech is now over 50 years old, and it is fascinating to see what a successful military commander elected president sees as worthwhile - schools, hospitals, homes, power plants, roads, without making a distinction of which of those are funded by private investment or public expenditure.

Only one quibble with the your last sentence. All expenditure is private. Public expenditure is appropriated from the private. Why would there be a distinction?

All public spending in all private income.

Sorry, you can't argue that government only takes from the private - every penny government takes from the private sector is returned to the private sector.

Every penny capital takes from labor must be returned to labor, but the problem in the past three decades is capital returns a large portion of what it takes in the form of debt that labor will never be able to repay. Even if capital pays labor lots more in higher wages.

Capital has lent labor directly and through public debt so much to consume that capital would need to pay labor more than what labor produces so capital redeems its debt faster by losing money faster on sales of production.

Unless the conservative-approved redistribution of wealth is invoked: bankruptcy judges as big government technocrats deciding who the winners and losers are, deciding that the deep in debt people are the winners by transferring the wealth of capital to the people in debt forgiveness.

Of course, the past three decades have involved making workers the ones who capitalize capital with their savings, savings they will never get back if the redistribution of wealth by bankruptcy occurs.

In Japan, they have a fairly closed economy where they avoided tax hikes to fund rebuilding after disasters by borrowing from the people instead of taxing them, so bankruptcy would be in effect retroactive taxes on wealth.

For the US, bankruptcy would be retroactive taxes on past profits from selling to the US in addition to taxes on worker savings for retirement. Detroit did not pay high wages to government workers and did not pay the Federal government for SS, so it borrowed the retirement savings they promised the workers instead of taxing. Now the solution is to redistribute the savings of workers as retroactive taxes.

Also, this is Britain 1934. Defense spending would very much offset hospital spending in the coming decade.

I think one thing we can say for 1934 Britain is that the overwhelming majority of the pennies spent on war production were well worth it.

"Every gun that is made, every warship launched, every rocket fired signifies, in the final sense, a theft from those who hunger and are not fed, those who are cold and are not clothed."

So, the food is taken from the mouths of the homeless and hungry who have no job, income, home, and are starving, in order to pay people to work to build guns and as a side effect means they can buy food, housing?

Instead of spending on guns, there should be no spending, no jobs, and the homeless and hungry with no incomes will have full bellies from eating what???

Why can't the laws of nature be constraints on economics?

Why is it that it that economists simply get to say that by being removed from the economy you are now no longer in need of food, housing, clothing, etc?

Im reluctant to engage you at all, but:

"Instead of spending on guns, there should be no spending, no jobs, and the homeless and hungry with no incomes will have full bellies from eating what???"

This is just reading fail. The speech prior quoted is pointing out that military spending comes at the expense of other, perhaps more preferable types of spending. Your trotting out a strawman of 'guns or nothing'.

I will wager that you never read the rest of Eisenhower's farewell address. He specifically warned about the scientific establishment.

Key passage:

Nevertheless, our best-guess impact is actually quite similar to the £326.4 million increase in output implied by Thomas’s approach on the basis of a multiplier of 1.6 times the additional defense expenditure of £204 million in 1938 compared with 1934. The difference is, of course, that our estimate is based on a much smaller multiplier, 0.52, but takes into account the news of very large defense spending plans for the future. The main implication of our analysis is that the observation of a large increase in output at a time of increased defense expenditure should probably not be interpreted as a large Keynesian multiplier but as largely reflecting firms gearing up for anticipated future increases in government spending.

As I see it, the problem with this argument is that the permanent vs transitory nature of defense news is a time series entirely invented by the authors - a lot of it relies on the interpretation of how businesses interpreted the British government's military intentions across the highly volatile interwar years. By admission, it relies intimately on, of all things, editorial stances taken by The Economist. The problem is that in the 1930s, The Economist was a highly niche magazine with a weekly circulation of maybe seven thousand. It was not a newspaper of record, nor a reliable podium to the business community.

@david - seems your quibbles with the data are misguided, see below. Also it's important to realize this Keynesian multiplier is for rearmament, not the general economy.

Ray Lopez

"The statistical information obtained from The Economist has been cross-checked against the detailed descriptions of British budgets provided by Bernard Mallet and Oswald George and by Basil Sabine.15 Interpretation of the commentary of The Economist has been facilitated by the accounts in the major historical studies of military policy such as those of John Ferris and George Peden.16 The general pattern of the news is quite clear but, of course, there is a margin of error in the details. At times, there was considerable uncertainty, not simply for agents in the private sector but also among policymakers, as to what would happen both with regard to magnitudes and timing, especially in the early 1920s and the late 1930s, and judgment calls are unavoidable. Expected values have been calculated at 1938 prices for a horizon of five years using a discount rate of 5.1 per cent.17"

Yes, but the defense budgets are not the point. As the authors themselves discovered, their non-Keynesian-multiplier "impact" factor is about the same as how the old Keynesians estimated their expectationless multiplier. The important bit is the author's estimates of expectations - the commentary in the magazine, not the statistics. That's what separates short-run multipliers from anticipation.

Correlating it with the studies of postwar historians is completely meaningless because of course postwar historians will tend to correctly anticipate military changes, they've already lived through them! Are we expected to believe that the authors, armed with knowledge of what would eventually happen to British military policy, are interpreting an editorial written in 1935 with an entirely neutral eye?

I see. But your last sentence should read not "authors" but "post WWII economists", for you see the authors in fact are anticipating your objections and using contemporaneous Great Depression sources to calculate the Keynesian multiplier. That is the point of their paper (see the paragraph citing V. Ramey et al). And it should be noted that the authors are not ideologically "die hard Keynesians", see for example this quote: " Our estimate is that, in the absence of the rearmament program, real GDP in
1938 would have been about 7 per cent lower"... 7% is not the end of the world, which a diehard Keynensian would predict. All in all, a good paper IMO.

I'm saying that their use of contemporaneous sources is intrinsically flawed! If you think otherwise, you are still not understanding what I am saying. V. Ramey's own paper makes the same mistake.

The point of the paper is to finely pick out the degree to which future changes in defense spending were anticipated, and their time series of expectations in this regard is completely made up by the authors' interpretations of editorial columns . That is a fact. It also applies to Ramey, who merely uses Businessweek instead of The Economist: again, another magazine whose circulation was much smaller in the period under study. It is impossible to believe that any author who knows that World War II is in fact going to happen, would be an objective judge of the expectations expressed in the editorials of a then-niche magazines.

I guess the ungated link above is broken, this is another one: http://www.econ.ku.dk/mehr/calendar/seminars/mehr24102012/Rearmament_to_the_Rescue.pdf/

"the fragility of our knowledge in this area": by "this area" you mean Macroeconomics, I expect.


Yes, the economy is being distorted by drunks on Sit 'n Gos, with the expected results.

I favor more research, and more in-depth research, on the subject of Keynesian policies, but I fear researchers will find what they are looking for. Today's economic distress and remedies for it won't be found by comparing the 1930s, for in the 1930s, the government did not intervene when the financial system collapsed, unlike in 2008-09, and the economy collapsed along with the financial system. Original research today would focus on economic recovery following a financial collapse in which the government intervened. It would focus on the short and long term effects of intervention. To the mere observer, intervention avoided economic collapse but preserved the economic instability preceding it. Focusing on the 1930s to "prove" this or that about Keynesian policies is misplaced.

Seems you are introducing a sort of Heisenburg Uncertainty Principle to economics: that intervention by the government would change the outcome of the economy, which is essentially a Keynesian worldview (compare with Fischer Black, who believed the Fed follows the market, and thus not even short term does the Fed really influence the economy). As to your second point, about 'original research', the Rogoff et al authors of This Time Es Different did just that--they found that financial crises take about 20 years to iron out based on several dozen data points. Critics of Rogoff et al would disagree however that it is right to lump the USA, which controls its own currency, with a country like Bolivia or Zimbabwe or Greece.

You missed the point, and difference: the intervention I am referring to is saving the financial system, which did not happen in 1928-30 but did in 2008-10. The economy followed the financial system and collapsed in 1928-30, and followed the financial system in 2008-10 and did not collapse. A purest would observe that the ups and downs of the economy must run their course, and would attribute the long period of shared prosperity after the 1930s to letting the financial system along with the economy collapse in the 1930w, whereas the intervention in 2008-10 just preserves everything that is wrong with the current economy.

I see, but essentially your argument is Keynesian if you think about it: you think the Fed or government can somehow 'save' an economy, or 'not save' it, depending on whether they do things like fiscal stimulus (pure Keynes) or monetary stimulus (supposedly Milt Friedman-ish different, but cut from the same cloth and the flip side of the same coin). My point is that Dr. F. Black may be right about not just the Black-Scholes options formulae, but also that the Fed don't matter long term (nor does Keynesianism). Might the Great Depression simply have been the result of an economy adjusting from the age of steam and horses and non-unionized labor to the age of electricity and motorized vehicles and unions?

I'm suspicious of any multiplier that is a positive number.

All multipliers can't be non-positive.

From page 11 of Nicholas Crafts and Terence Mills:

"...So where did recovery come from? Insofar as it was stimulated by policy, the initial phase was based on leaving the gold standard and ‘cheap money’. The policy stance was developed quite fully by late-1932. It entailed low nominal interest rates and a commitment to raising the price level, underpinned by an exchange-rate target of a 25 per cent nominal devaluation compared with the gold-standard parity, which was enforced through intervention in the foreign exchange market.53 In terms of the overall expansion of real GDP (25.8 percent between 1932Q2 and 1938Q4), our estimate indicates a contribution from rearmament of about 30 per cent, building up mainly after 1935.54..."

An examination of Table 5 reveals that defense spending did not increase significantly until 1935Q4. The RGDP data comes from Table 2b of Mitchell and Solomou (2009):


Between 1932Q2 and 1935Q3, a period of 15 quarters, RGDP increased by 14.4%. Between 1935Q3 and 1938Q4, a period also of 15 quarters, RGDP increased by 9.7%. In other words the rate of RGDP growth decreased *after* the defense buildup commenced.

Hmmm. Sounds like yet another example of monetary policy offsetting fiscal policy at the zero lower bound.

"Oh where, oh where has the liquidity trap gone?
Oh where, oh where can it be?
With its interest rates low, and fiscal multipliers high,
Oh where, oh where can it be?"

Excuse me, 13 quarters.

This was a difficult paper for me, partly because the Multiplier Effect, to the extent that I see the usefulness of it, would be heavily Contextual, depending upon the specific economic effects you want to counterbalance at that time. So, when the authors say this:

"The latter shows that defense news has a negative effect on non-government spending which is indicative of crowding out."

I don't quite see how that follows, unless Crowding Out means something other than taking money away from investors that were on the verge of investing, and transferring the money to the government for investing. This is 1930s England. If my dream is to open up a kosher butcher shop in London, I might be having second thoughts. And, indeed, the authors then say this:

"As Roger Middleton has stressed, contemporaries, including key officials in HM Treasury, thought in terms of ‘psychological crowding out’, in the context of deficit-financed public works programs, through adverse effects on business confidence which might undermine investment, as a major reason to believe that the multiplier effects would be small."

But this conclusion seems backwards. The Loss Of Confidence is from the possibility of impending hostilities, which is why Defense Spending is growing. And then this:

"Middleton also noted that, in the debate on the budget in 1933, the Treasury publicly maintained that any possible expansionary effects from an unbalanced budget might be vitiated by expectations of future tax increases and that the strong public commitment to the balanced budget rule by government ministers meant that any suggestion of a deficit would lead to expectations of higher taxation."

Future Taxes? In 1933 England I'm just hoping to be alive to pay taxes. Did anyone during The Blitz really worry mainly about the Public Debt and Future Taxes?

The Multiplier Effect, I thought, was about Govt Borrowing/Investment increasing GDP in a time period when Private Investors are loathe to invest. As for Predicting Future Tax Burdens, I don't see how you can reasonably predict this amount based only upon Servicing the Debt. There is no A Priori Reason that Govt Spending and, hence, Tax Burdens, need to Increase after a Financial Crisis. You must need to be assuming that, in fact, the problem will be that Taxpayers will balk at paying the necessary future taxes, which is the opposite of blithely assuming a tax increase.

Having said all this, some of their conclusions do seem at least partially explanatory, and I'm not sure I even find the notion of Crowding Out or Multiplier Effect all that useful. But, as I said, it was an interesting, but difficult, paper for me.

@DP - you seem to have lost your rhetorical flair when you state: "Future Taxes? In 1933 England I’m just hoping to be alive to pay taxes. Did anyone during The Blitz really worry mainly about the Public Debt and Future Taxes? " since you apparently are confusing 1943 with 1933. As I said upstream, the paper seems well reasoned, contrary to what David is saying, and Fair and Balanced. It's actually a point for the anti-Keynesian crowd more so than the Keynesians, which wins my heart.

No, Ray, you missed my point, which is my fault by overstepping my rhetorical bounds and mentioning The Blitz, which was, indeed, later. To follow my reasoning, you have to ask yourself why defense expenditures were increasing from 1933 on, say, in Britain. I think it has to do with what was going on in Germany, which, had I been alive then, would have scared the hell out of me, and, to a large degree, colored my view of investment. As the 30s rolled on into the 40s, any concern I would have had with the public debt would have fallen away. That was my point about the Blitz. I'm saying this is a very specific historical circumstance which might have specific effects on any Multiplier, if there is one.

As for Keynesianism, I don't really feel the need to get into that dispute.

Comments for this post are closed