Now that the U.S. government shutdown is behind us it is instructive to review what we went without for those weeks. Recently writing for the Financial Times, Robin Harding reported that experts feared that the U.S. government would be “flying blind” without access to the big data collected by the government, and which was imperiled by the shutdown.
We would have been so lucky.
Big data does not provide the necessary information for government officials to “steer” the economy. It only serves to give the appearance that someone at the top is in command of the big picture, and altering the course of the economy in such a way that the big numbers newspapers covet so much – the unemployment rate, inflation or the omnipresent GDP – are aligned with the prosperity voters expect.
Government officials are already flying blind, with or without big data. The reason is twofold and simple: the data necessary for business decisions is not often available in an easily recordable form, and even if it was, the relevant information is never of the general form that government uses but of a localized and very specific nature.
Consider a businessman, perhaps a small tailor near Waterloo, Ontario. Our tailor needs to know what styles of clothes people are demanding, the fabrics they desire, and the weight of the cloth that will be necessary to get them through the coming winter. None of these criteria are in an easily transferable form, indeed, many of them are fundamentally tacit. If asked why he purchased a certain sweater to stock for the winter he probably cannot give a quick answer. The tailor just “knows”, whether by gut or instinct or by whatever term one calls entrepreneurial foresight.
Our tailor cares not how much general inflation rate in Canada is, or even the rate in Ontario. Indeed, he doesn’t even care much about what prices in general are doing in Waterloo. He is very much concerned with what the incomes of his customers are doing, where the prices of competitors’ clothes are going, and what how much his expenses will fluctuate by. None of these data are available through the general government’s surveys.
Business conditions are never the result of general changes permeating through the whole economy. The causality is the other way ‘round.
Businessmen use information specific to both time and place to alter their goods and services on offer. Consumers do the same thing – surely you don’t care what the average temperature in Canada is expected to be this December, just as you don’t care what the average rate of inflation will be next year. You care about those factors that matter to your immediate vicinity: the relevant data will always be of a localized nature. The interaction between consumers and producers, each acting upon their own little local slice of the world of information, are what create the general figures that the government compiles and tries to pass off as important.
Thus all the individual sweater sales and purchases interact to create some contribution to the larger GDP and inflation figures for all of Canada, but the latter two figures have almost no bearing on the individual purchaser’s or seller’s original decision.
If anything, big data serves a purpose of giving a false legitimacy to government policies. The policies that the government enacts are always of a necessarily broad nature – that is the inevitable result of having a federal government. Perhaps a less strong statement would just point out that the relevant trade off is between the size (level) of government and the generality of its policies. The more general a policy the less likely it is to have a meaningful effect of individual economic actors within the system.
Consider that the Bank of Canada is worried that inflation is low right now – big data tells it so. The Bank’s governor, Stephen Poloz, responds by increasing the money and credit supply in a bid to push prices up. But the real problem plaguing Canadian businesses right now is that some prices are too high (and these businesses are uncompetitive) and some prices are too low (and these companies are not making adequate profits). The agricultural sector is a good example where there are both problems at play. The Bank of Canada’s blanket policy has increased all prices, thus aiding one group but not the other.
In this example big data enabled the BoC to pursue a misguided policy. It furthermore gave it an aura of legitimacy by making it seem as though its goal, increasing inflation, was both relevant and important to Canadians. It is not, though specific prices over which the BoC has almost no control are most certainly relevant and important. We – businesses and consumers – create and alter these prices through our own actions, not those of the government.
Instead of lamenting the potential loss of “big data” by the U.S. government shutdown, we should be rejoicing. Not only will it allow us the opportunity to see that the economy can do just fine without it, but it may also give us a break from the misguided economic policies structured around it.