This Thursday the Conference Board, a global business association, released its monthly index of “leading economic indicators.” Like the unemployment and inflation, housing starts, G.D.P. changes and other figures, these numbers arrive in metronomic waves. Financial services like Bloomberg, Dow Jones and Reuters blast them out the moment they’re released. Stock markets will often respond within seconds. Commentators and policy makers attribute to them a near-cosmic significance.
We act as if they are markers from time immemorial, but in fact they were invented for modern industrial nations after the Depression and World War II and are now seriously outdated.
Take gross domestic product. Derived from formulas set down by the economist Simon Kuznets and others in the 1930s, its limitations have long been recognized, none more eloquently than by Robert F. Kennedy in a famous speech in 1968 when he declared that it measured everything except that which is worth measuring.
GDP treats all output as a positive. When you buy LED lights that obviate the need to spend on incandescent bulbs and reduce energy consumption, GDP goes down and what should be an unmitigated good becomes a statistical negative. If a coal company pollutes a river, the cleanup costs are positive for GDP, as are any health care costs for those harmed.
What’s more, we have also come to assume that with output comes more spending and employment, but factories today are powered by robotics and software, and robots don’t buy more lattes and shoes.
GDP is a good number for a nation that produces lots of stuff made by lots of workers, but for an information economy grounded in services and intellectual property and awash in apps that cost nothing yet enable commerce, it is not up to the task. Nor are many of our indicators. Our trade figures treat an iPhone made—more accurately, assembled—in China with no reference to the intellectual property created by Apple in California.
Our inflation numbers can barely keep up with the quality improvements embedded in our cars and appliances and homes, despite heroic efforts by statisticians to account for those efficiencies in their formulas.
Our employment numbers are mute about 20-somethings who leave their paying jobs to create new ones.
And our national figures treat each “economy” as a sovereign island—even though flows of labor, goods and services are as indifferent to national boundaries as weather. Our lives may be ever more global, but our indicators remain stubbornly national.
We would be better served by confronting the vagaries of the world around us, unclouded by a statistical construct of an earlier time. Economic policymakers have need of some macro indicators, but few others do. The invention of the economy helped tame some of the worst of business cycles in the 20th century, but increasingly our indicators constrain our ability to understand the world and then act creatively to shape it.
There have been multiple efforts to create newer, better indicators. Most notable in recent years, Bhutan announced a gross happiness index. Less self-servingly, the United Nations developed a Human Development Index. Others have proposed new ways of assessing just about everything. All of these, however, still reduce complicated systems to simple, round numbers, and all of them therefore will fail to describe our world, even if they fail for different reasons.
What we need to do instead is access the information revolution to craft our own bespoke indicators. We are now deluged with data, and we no longer need simple one-size-fits-all national indicators. Instead, we need to use big data to fashion solutions. The result will be a plethora of bespoke indicators, each tailored to particular needs.
For individuals, national economic numbers have little direct relevance. Buying a home, getting a degree, seeking a job, deciding on how much debt to take on — none of those decisions are meaningfully shaped by GDP growth or aggregate inflation and housing numbers. Deciding whether to start a small business is also marginally influenced by such statistics: local trends in a particular industry or geography matter more.
In Nebraska, for instance, there has been no unemployment crisis over the past five years, and with high grain prices and the shale oil revolution, localities are doing just fine. Contrast that with Detroit, central Florida or the exurbs of Los Angeles and Phoenix.
Yes, large corporations have economists who attempt to draw correlations between macro-indicators and business trends, and companies decide on how to much to spend based in part on a read of future interest rates, growth trends, and inflation. But even here, the connection between big numbers and business realities has broken down. If national retail sales that measure big stores in malls are weak, that says nothing about how much e-commerce might be up. If consumer spending writ large sags, that says nothing about higher end spending at mass luxury stores like Michael Kors or lower-end retailers such as Dollar Tree. Making decisions based on what the indicators say is almost certainly a recipe for making the wrong decisions.
Of course, policy makers need some reference to numbers, which feed into their complicated statistical models of “the economy” and help the Federal Reserve to set monetary policy (and the White House and Congress to set fiscal policy, if they can manage to agree on it).
Even here, however, officials are beginning to recognize that steering policy by them is not wise. The new Fed chairwoman, Janet L. Yellen, recently acknowledged that the headline unemployment rate was less and less useful for policy-makers at the central bank because the figure doesn’t capture any of the nuances of underemployment, discouraged workers who have stopped looking, dropped out, freelancers and interns, and the long-term unemployed.
Similarly, policies meant to juice consumption and boost GDP might help game the numbers, but they do little to enhance the long-term viability of a dynamic economic system. The stimulus bill passed almost exactly five years ago promised that with $800 billion in spending would come 3.5 million jobs. That never quite happened, because the correlation between spending and private sector hiring, while stronger in 1950s and 1970s, is no longer as potent in an information economy where cheap capital, labor and technology make it easy for companies to generate massive profits and extract more productivity from their workers without having to hire more of them.
Weaning ourselves from our obsession with economic indicators is a vital step to grappling with the world as it is and making decisions that yield positive results. Individuals, companies, and governments will find their interests best served by creative approaches that craft indicators that draw on the wealth of big data information rather than cramming all reality into a few simple averages. The indicators of the 20th did yeoman service in taming the worst extremes of economic cycles. We should thank them, and move on.
This post is adapted from Zachary Karabell's The Leading Indicators.