Britain’s Central Statistical Office (now the Office for National Statistics) was founded in 1941 following a direct order from Winston Churchill. With Nazi troops at channel ports, the issue to be answered was obvious. What was the capacity of the British economy to build tanks, guns and planes? And how successful was wartime direction of production in diverting resources to these uses?
Both the relevant questions, and the nature of the economy, are today very different. But the methods by which national economic statistics are collected have changed less, as last week’s Treasury report on the subject from Sir Charlie Bean shows.
Greater quantities of physical goods are less and less important to these expectations. An affluent society mostly wants better, rather than more. Increasingly, it wants services rather than manufactured goods. Goods and services come in infinite varieties and many qualities. How can I even interpret a measure of output of goods, when the phone in my pocket has more processing power than the massive university computer on which I did research in the 1970s?
Most of these goods and services are not valuable in themselves, but only as means to some end. Of course, this was always true: Churchill did not really want Spitfires and Hurricanes, but things that would repel attacking Germans. Today, I do not really want a phone in my pocket — to remember and carry it is a nuisance — but I do want to be able to talk to friends and colleagues, read books, find my way from A to B and share photographic memories. The genius of the smartphone is that it does all these things.
What we need to know today in order to understand the effectiveness of a modern economy in fulfilling the aspirations of modern consumers is fundamentally different from what we needed to know in order to direct wartime production effectively. Yet the picture Sir Charlie describes is one in which, to a considerable extent, official statisticians are just turning the same handle that Churchill ordered them to turn 75 years ago.
The report contrasts the extraordinary granularity of the categorisation of manufactured output with the broad brush classification of categories of services. We are slow to acknowledge new methods of production and consumption; Sir Charlie notes that the omission from official statistics of the room-letting website Airbnb alone might reduce reported gross domestic product by as much as 0.7 per cent.
Depreciation of capital is secondary when there is a war to be fought, but central to the quality of modern life. Besides these fundamental questions, the matter of whether revisions to provisional data estimates in Britain are larger or smaller than those made by statistical offices in other countries seems inconsequential.
In 1940, the only way to assess the productive capacity of the British economy was to survey plants and ask their managers what they could do. But now the information technology revolution generates extraordinary quantities of information every day. Computers and cameras are recording every sale, every movement of goods and every transaction of significant value.
Everyone is talking about “big data”. Does it not seem odd that the principal means of compiling official statistics is still the specially commissioned survey?
This article was first published in the Financial Times on March 16th, 2016.