The world of business television, full of scrolling quotes and zigzagging graphs, is an illusion. It tempts us to believe that we understand what is happening in the economy, as it happens. We do not.
The share price of Toyota, for example, reflects today's supply and demand for Toyota shares - not supply and demand for cars, which we cannot observe. House price indices purport to measure how the price of 'houses' has changed over the past month. But the houses bought and sold last month are not the same as the houses bought and sold the month before. It is not easy to establish whether this month's houses are more expensive because, in some abstract way, housing itself is more expensive or because estate agents sold a couple of mansions in Belgravia.
The most important measure of economic activity, GDP, arrives in fits and starts, with early estimates giving way over time to more informed measurements. Forget economic forecasting - we don't even understand what is happening now. Will this ever change? Maybe, but not without a rethinking of where we get our economic data from. Computer scientists are now wondering whether data already flowing through digital networks may help us to understand apparently unrelated economic activity.
A quirky example was published last year: "Twitter mood predicts the stock market". Researchers searched Twitter for expressions of mood such as "I feel stressed", and found that when a sense of calm descended on the Twittersphere, it could be used to predict market movements two or three days hence.
I am doubtful this is a lasting empirical regularity but other computer-based indicators may prove more useful and less transient. Take searches on Google: while the search giant knows a lot about us and keeps much of that information to itself, it does release data on overall search volumes for any search term you care to name.
The Bank of England is paying attention to how many people in the UK search for 'estate agents', for instance, or 'JSA' (Jobseeker's Allowance). This is revealing: searches for 'JSA', for example, are correlated with official unemployment data but are available a month earlier. It's not quite a peek into the future for the Bank's Monetary Policy Committee, but it is at least a peek at the present - which is no small thing.
If I was asked to speculate about the future of such computer-based 'nowcasting', I'd point to a recent research paper that uses data from a new system designed to collect road tolls in Germany, which tracks the itinerary of heavy vehicles. The resulting 'toll index' gives an excellent snapshot of German industrial production. Perhaps we can look forward to the day when real economic data is collected and disseminated almost as easily as the charts on business TV.
blog comments powered by
