I’ve been re-reading Robert Skidelsky’s original biography of John Maynard Keynes (The [amazon_link id=”0142001678″ target=”_blank” ]Fighting for Britain 1937-46[/amazon_link] volume, not the recent [amazon_link id=”0141043601″ target=”_blank” ]Return of the Master[/amazon_link]). In particular, refreshing my more than 10-year-old memory of the origin of national income accounting at Keynes’s instigation. Since 1937, Keynes had been urging the government to collect the appropriate statistics. He wrote: “Every government since the last war has …regarded the collection of essential facts as a waste of money.”

But, as Skidelsky continues:

“What he omitted to point out was that such facts were essential only for the purposes of ‘demand management’ which the pre-war Treasury regarded as neither feasible nor desirable.”

According to Roy Harrod’s earlier (1951) [amazon_link id=”0393300242″ target=”_blank” ]Life of John Maynard Keynes[/amazon_link], soon after publication of [amazon_link id=”B005E8FUT4″ target=”_blank” ]How to Pay for the War[/amazon_link] in 1939 had introduced the concept to a wider public, Keynes circulated privately a pamphlet A Budget for National Resources. Subsequently Austin Robinson at the Treasury set Richard Stone and James Meade to work producing the UK’s first national accounts. This was published with the 1941 Budget. According to Harrod, it was: “A great revolution… This kind of accounting has come to be regarded as the essential tool of any economic planning, whether of an Individualist or a Socialist variety.”

In their superb recent book on Keynes, [amazon_link id=”0674057759″ target=”_blank” ]Capitalist Revolutionary[/amazon_link], Roger Backhouse and Bradley Bateman note that Paul Samuelson, reflecting on the parallel work on national accounting in the US described the war as , not the physicists’ war, but ‘The economists’ war.’ They say: “Economics emerged from the Second World War with its reputation greatly enhanced.”

It is an interesting perspective on the current debate about measuring the economy. GDP was never intended as a measure of welfare or well-being, but as a measure of the nation’s productive capacity in wartime, for the purposes of government planning. The fashionable survey measures of ‘well-being’ are obviously a rubbish indicator for policy purposes. There is, though, a question about whether GDP, adapted so much over the years in successive new international standards, remains the right indicator for government policy. What war should economists be fighting now?

[amazon_image id=”0142001678″ link=”true” target=”_blank” size=”medium” ]John Maynard Keynes. Volume Three: Fighting for Freedom, 1937-1946: 003[/amazon_image]

I’ve been reading this morning Richard Stone’s 1948/49 Newmarch Lectures, [amazon_link id=”0751201863″ target=”_blank” ]The Role of Measurement in Economics[/amazon_link]. There really, truly is nothing new under the sun. He writes in the conclusion about the ‘undue exclusiveness sometimes to be observed in economists’:

“The determination to deal only with the economic aspect of problems doubtless makes for tidy theories and all the pleasures of staying on home ground. But inasmuch as the economic aspect is only part of most actual problems, its single minded pursuit leads to distortion and incompleteness. This is particularly so at the present time when social influences are again coming to play a larger part in shaping the changes in society than they did in past eras… The moral seems to be not less specialization but more co-operation and understanding between economists and those who work in other branches of the social sciences.”

He also – as could be expected from such a leading figure in developing the modern system of national accounts – advocates the collection of adequate statistics for empirical understanding. (And the need to educate the general public in some fundamentals of statistics so that everyone understands there is a margin of error.)

Lastly, Stone strongly urges the need to combine empirical and theoretical approaches to economics. Theory versus measurement is a false dichotomy, and theory without measurement uninteresting.

British Economy, 1948

I just read for the first time since the late 1970s Keynes’s essay [amazon_link id=”B005E8FUT4″ target=”_blank” ]How to Pay for the War[/amazon_link], published in February 1940 on the basis of two Times articles of November 1939. (It’s included in my 1972 RES/Macmillan edition of [amazon_link id=”0230249574″ target=”_blank” ]Essays in Persuasion[/amazon_link].)

My immediate interest was the part this essay played in the development of official national output figures in order to measure the economy’s productive capacity: Keynes’s frustration at needing to rely on the private estimates recently published by [amazon_link id=”0714612162″ target=”_blank” ]Colin Clark[/amazon_link] is evident.

[amazon_image id=”0714612162″ link=”true” target=”_blank” size=”medium” ]National Income and Outlay[/amazon_image]

However, I’m struck – awed – by the clarity of Keynes’s overview of the way the war economy needed to be managed. The ability to step back from small parts of a problem to see the whole is rare. In this case, it involved understanding that a different way of thinking was required, so greatly had the context changed. Keynes wrote:

“We have become so accustomed to the problem of unemployment and excess resources that it requires some elasticity of mind to adapt our behaviour to the problem of full employment and of resources which are no longer adequate to supply our needs.”

Elasticity of mind – a lovely phrase for a capacity so few of us have.

This question of changed context is surely relevant to today’s macro policy questions. It often seems to me that the dispute between different camps is a matter of disagreement about context. In other words, they disagree about the nature of the most pressing problem. I’ve always been very taken with the way Edmond Malinvaud set this out explicitly in [amazon_link id=”063117690X” target=”_blank” ]The Theory of Unemployment Reconsidered[/amazon_link], although I don’t think his specific ’70s framework fits the kind of open economy and financialised world we have now.

[amazon_image id=”B005E8FUT4″ link=”true” target=”_blank” size=”medium” ]How to pay for the war : a radical plan for the chancellor of the exchequer / by John Maynard Keynes[/amazon_image]

In the course of my research yesterday, I came across this from Colin Clark, in his 1940 book [amazon_link id=”B0006DB6N6″ target=”_blank” ]The Conditions of Economic Progress[/amazon_link] (via an Angus Maddison essay on Clark):

“There is room for two or three economic theorists in each generation, not more. Only men of transcendental powers of reasoning can be candidates for these positions. The rest of us should be economic scientists, content steadily to lay stone on stone in building the structure of organised knowledge”.

The scientific approach he defined as: “Careful systematisation of all observed facts, the framing of hypotheses from these facts, prediction of fresh conclusions on the basis of these hypotheses, and the testing of these conclusions against further observed facts.”

To put this another way, he argued for more inductive than deductive thinking in economics. We’ve had it the other way round, I fear.

The Conditions of Economic Progress

This story about the Public Accounts Committee‘s call for public data to be made easier to use sent me to two of my favourite reference books, Edward Tufte’s [amazon_link id=”0961392142″ target=”_blank” ]The Visual Display of Quantitative Information[/amazon_link] and Howard Wainer’s [amazon_link id=”0691134057″ target=”_blank” ]Graphic Discovery: A Trout in the Milk and Other Visual Adventures[/amazon_link]. There is an apt Tufte quotation: “Good information design is clear thinking made visible, while bad design is stupidity in action.” This is just as true of words: if you don’t think clearly, you can’t write clearly. Writing that is not clear might not reflect a lack of authorial understanding, but I tend to take it as a serious possibility.

[amazon_image id=”0691134057″ link=”true” target=”_blank” size=”medium” ]Graphic Discovery: A Trout in the Milk and Other Visual Adventures[/amazon_image]

When it comes to data and the presentation of statistical analyses, what the PAC says is obviously correct in a way: true transparency involves structuring data, not sending it out in a raw torrent. On the other hand, structuring data requires a lot of time and effort, and public agencies have neither the resource nor necessarily the expertise. So beyond a certain point it is surely fine to expect somebody else to make sense of the data and present it appropriately.

However, there is surely still a gap in the market for very easy to use (and affordable) software to handle data and present it flexibly. Part of the tyranny of Powerpoint, against which Tufte rails so powerfully, is the rigidity of the graphing tools. I know there are lots of other visualisation tools available and The Guardian for one has done terrific work educating us about this, but that’s the challenge – there really are lots. Here’s another list and another of data visualization tools and another. Here’s Tableau’s free software. Lots of it out there. But nothing has become the standard to replace Powerpoint when you’re in a hurry. This is the path dependency/first mover effect in action. OK, one has to invest the time in learning new tools, but I haven’t yet found one to settle on, and don’t want to learn fifteen. No doubt somebody will tell me what an idiot I am for not having discovered the answer to my problem – I hope so!

[amazon_image id=”0961392142″ link=”true” target=”_blank” size=”medium” ]The Visual Display of Quantitative Information[/amazon_image]