Feed on
Posts
Comments

More or Less Data?

As a "professional" economist it is natural that I would seem to support efforts by the government to collect data. These function very much like employment subsidies for me. Do I think that the collection of data is subject to "public goods" problems? In some respects yes – not that data cannot be excluded from non-payers, but that some valuable data may simply not be collected for the same reason that you may expect private industry to "under"fund basic scientific research.

Having lots of data would seem to be a great boon to us economists – allowing us to build elaborate macro-econometric models, estimate how significant discrimination is, estimate how much we value a (statistical) life, and much-much more. On the other hand, think of what having access to lots of data also does. It allows planners to have a great big tool to use to plan the economy. It's been very entertaining to read the debate among Monetary Theorists over the last 4 years about whether NGDP has been growing fast enough and whether employment has responded to changes in NGDP and real wage cuts the way our models suggest they ought to. It has been entertaining to read about the supposed impacts of stimulus. It has been interesting to read about changes in our trade balance as the global economy has teetered. But what if none of this data were available? Would it get created? Do we think that the planners would have fewer tools at their disposal and hence have less of a tendency to plan?

I don't know the answer. I used to think that having less data would make us less inclined to plan, and to force ourselves to revert to thinking about sound fundamentals when it comes to making policy, but I really don't know. What do you think?

3 Responses to “More or Less Data?”

  1. cmprostreet says:

    Interesting question.  I'd be afraid that if planner's can't have data to inform their schemes, they'll start planning by anecdote.  That does have the potential to be even more entertaining, though.

  2. Harry says:

    What a great question, WC. I know you think like Hayek, but without getting into the many avenues you have invited us to go here, let me make a micro comment from my experience.
     
    In the macro world in which all Wall Street types swim, there is a closely-wathced number, capacity utilization. When that figure is published periodically, every dutiful CFA plugs that figure in, along with the rest of the numbers he/she gleans from government and other sources, into their computer model.
     
    Having lived in the productivity/operations consulting world for many years, I do not think I ever once walked into any business where my client had the slightest idea of what their capacity was. Indeed, that was why we were in business — to improve the ratio of output over input, the definition of capacity and productivity.
     
    Everywhere we went, we found loose standards for work and quality — departments operating at 100+ percent of standard generating scrap and rejected material, if it was caught before it left the shipping floor. Too many people, too much equipment, too much space, too much inventory, too much of everything everywhere, all the time, including too much time.
     
    Now, periodically some of these businesses, at least the big ones, reported figures to the government about capacity utilization. I can guarantee that every number sent to the government was close to imaginary. I would not trust that number to make an investment decision about my own money for a second, even if it appeared in Barron's Market Laboratory. Not in the model in my head.
     
    I have found the Purchasing Managers Index, the price of gold, and other numbers useful. Every number, however, is a function of what happened yesterday, after the accounting department has massaged it. Sales? Depends on how many units have left shipping and loaded onto rail cars or into trailers.
     
    Today in the WSJ Arthur Laffer talked about how businesses have probably already moved profits into 2012, in anticipation of Taxmageddon. (The problem is not uncertainty; if they were to make Taxmageddon certain, then everybody would run for the door.) Plug that one into your macro model.
     
    It really scares me when Princeton professors, smoking their pipes in the salon, discuss how to guide the ship of state, confidently using imaginary numbers.

  3. Michael says:

    I think that by taking away government provision of data, it would make the price of doing research more expensive, but I'd bet the quality of the data itself would go up.  At least we would hopefully stop looking at the balance of trade, which I consider to be the most useless statistic commonly used. 

Leave a Reply to cmprostreet