David R. Henderson  

More Data or Less?

The Center Will Hold: My 10:1 ... The Fed is above the law...

My friend and Hoover colleague John Cochrane has signed a letter to four members of Congress that calls on Congress to spend more money on getting better economic data. John argues at his blog:

Few public goods are as cheap or important as good economic data. Much of our national policy discussion is based on government-collected data. Changes in inequality, wage growth or stagnation, employment and unemployment, growth, inflation... none of these are readily visible walking down the street.

Free, openly accessible, well-documented data, allowing comparisons over long periods of time, such as provided by the Bureau of Labor Statistics, is especially valuable.

That's all true. But the big question is: how will these data be used?

Another Hoover friend and colleague, Alvin Rabushka, writes:

What are economists likely to do with better data if their letter succeeds?

My guess is to advocate more funding to fight poverty, not more efficient use of existing appropriations.

More government funding to support greater participation in labor markets, not removing regulations, high taxes, and other impediments to labor market participation.

More stimulus to increase GDP, not less government interference in the economy.

More redistribution to reduce income inequality.

Here's someone else at Hoover who seems to agree, at least in part, with Alvin. In a moving story about a business hounded out of existence by government regulation, he writes:
The anecdote has two larger implications, in my view. First, many people including myself have a sense that this kind of regulatory persecution is harming economic growth. But there are no good estimates of the total number of companies put out of business, jobs destroyed, investment made worthless by regulatory persecution, or, even harder, projects not started from fear of such persecution. For the moment, we can only collect anecdotes, which is why I pass this one on. But the measurement question is vital. Inequality is only a big policy issue because we have statistics on it.

Here, this Hoover colleague makes a case both for more data, presumably collected by government, and, given his views on inequality, less government data. Who is this Hoover colleague? The above mentioned John Cochrane.

COMMENTS (5 to date)
BorrowedUsername writes:

Sounds like a really weak argument to me that we shouldn't collect data because some people might use it for prescriptions you disagree with. Even if your right the solution is to improve the statistics, not fail to collect them. It's worth making sure the process of what is / isn't gathered for data isn't overly politicized, but then again, if you knew that a certain piece of data prefers one side of the argument over the other, make sure you collect the data that puts it into context.

JLV writes:

Basically, data might prove your viewpoint wrong, so collecting data is bad.

That's many things, but honorable is not one of them.

MikeP writes:

Sounds like a really weak argument to me that we shouldn't collect data because some people might use it for prescriptions you disagree with.

That "really weak argument" is one of the most important factors behind the miracle that is Hong Kong.

From Marian Tupy's rememberance of John Cowperthwaite:

We spent hours talking about Hong Kong’s 16 percent tax rate, business-friendly regulatory environment, lack of state subsidies, tariff-free trade relations with the rest of the world and other policies he promoted while Financial Secretary. Of all the policies that we discussed, one stands out in my mind — if for no other reason than because it is so thoroughly counterintuitive. I asked him to name the one reform that he was most proud of. “I abolished the collection of statistics,” he replied. Sir John believed that statistics are dangerous, because they enable social engineers of all stripes to justify state intervention in the economy.

A "fact" as I understand it can only exist in the context of an interpretive theory. Didn't Uncle Freddy say as much in The Sensory Order? So any quest for data must reflect an underlying set of theories, a way of seeing the world. People with differing world views will differ in the data they want, because they want data to support their advocacy in the assumed mechanism of social change (public education then elections).

The good people who assemble the data on economic freedom of the world have to leap over some pretty big gaps, I suppose, as they attempt to quantify "impartial courts" or "reliability of police". Such variables are crucial underpinnings in my worldview. But should we look to governments as an unbiased source for such data?

One thing I think I've figured out is that government creates its own view, its own way of seeing what is important (in that view!) by assembling and making public data which supports the public-policy worldview. For example, I can probably get a pretty clear view from government data on population and voter registration in various counties in the US. But I do not expect government to be good at collecting data the purpose of which is to undermine the goodness of government.

In a society with property rights many things will be obscured from my view by those property rights. For example, in America it is still normally the right of parents to choose what to read to their children at bedtime. There is no collection of data of what my neighbors read to their children at bedtime. I am thereby blind in regard to my question about what my neighbors are reading to their children. A free society is largely a society of ignorance about what goes on within the realms of others.

David R. Henderson writes:

Basically, data might prove your viewpoint wrong, so collecting data is bad.
You badly misunderstood my point. I did not say that at all, nor would I, because I don’t believe it. Data that might prove my viewpoint wrong are always welcome. I don’t want to go around thinking things that aren’t true. It’s some of the uses that people put the data to that concern me.
Let me take an example. In 1942, the U.S. government used Census data to help round up Japanese Americans on the West Coast. That was a misuse of data. In that case, the more accurate the data, the better from the government’s viewpoint and the worse from my viewpoint.
See MikeP’s quote from John Cowperthwaite for the point I’m making.

Comments for this entry have been closed
Return to top