Abuse of Numbers

Sometimes what makes me want to blog is that I'm highly irritated by something I read. The danger is that it will make me seem bad-tempered. Oh well. Fareed Zakaria's article, "Obama's speech shifts focus from deficits" really irritated me. He begins:

With his speech in Kansas, President Obama has begun a national conversation about the economy and the role of government. In presenting his view, Obama shifted the economic conversation from deficits to the crucial issue of growth. After all, deficits matter because they could have a harmful effect on growth. The question we should all ask is: What would make this economy grow?

One theory heard a lot these days is that the economy is burdened by excessive government regulation, interference and taxes. All these pressures on business, especially small business, are keeping the economy down. Cut them, the Republican candidates all say, and the economy will be unleashed. It’s a compelling picture, but the data simply do not support it.

Now right here, already, the proper reaction to this is something like "Who does he think he is to be writing about numbers at all?" If you think a claim like this can be dismissed because "the data do not support it," then you don't understand the use of data. Data is not a substitute for theory. Theory is needed in order to interpret data. Disraeli said "there are three kinds of lies: lies, damn lies, and statistics." That's too cynical, but it's true that many, many uses of statistics are misleading to the point of being worthless, and people who think they can resolve big questions like this merely by quoting a few numbers are just the types of people who abuse statistics. Even so, I didn't think it would be as bad as it would turn out to be.

A World Economic Forum survey that ranks countries on their overall economic competitiveness puts the United States fifth; the countries ahead of it, including Singapore and Finland, are tiny, with populations around 5 percent that of the United States. The World Bank publishes a report that looks at “Doing Business” across the globe. The United States ranks No. 4, again behind a handful of tiny countries. As is the case with the World Economic Forum, that ranking has not changed much over the years.

Of course, there's plenty of controversy around the way these numbers are compiled, and the whole idea of "overall economic competitiveness" of countries is, in my considered judgment, baloney. To put it bluntly, it is totally eviscerated and refuted by Ricardo's theory of comparative advantage. But even if you take the number at face value, it does not support Zakaria's claim. The US has a light burden of government relative to other advanced countries, so it's quite "competitive." If it had a still lighter burden of government, the crude logic would suggest, we would be still more "competitive." What is Zakaria's point?

The Kauffman Foundation, which looks at the level of U.S. entre­pre­neur­ship, found that in 2010, 340 out of every 100,000 Americans started a business each month. That rate hasn’t changed much in the past few years; it is only slightly higher than in 2007, before the recession. Regarding regulations, Bloomberg News has crunched the numbers and found that the Obama administration has not reviewed or issued significantly more rules than its predecessors.

Again, what is Zakaria's point? That the pace of business development hasn't been stalled by over-regulation? But, first of all, while these raw numbers are interesting, they are hard to interpret. Not all rules are alike, after all; some are much more burdensome than others. And then, how are they enforced? That can change from administration to administration too. As for the number of Americans starting a business each month, we might expect that number to rise, since one reason for starting a business is to escape unemployment, and there are a lot more unemployed Americans now than in the past. Also, not all business expansion takes the form of individuals starting businesses; it can consist of established firms embarking on ambitious new projects. Really, not much can be drawn from these numbers.

The Organization for Economic Cooperation and Development (OECD) released a study last week measuring tax revenue as a percentage of gross domestic product. The United States came in 27th out of 30 countries. Taxes are low in historical terms as well, the lowest since the early 1950s. But the complexity of the U.S. tax code clearly exacts a price in terms of economic growth and competitiveness. The World Bank study finds that the only category in which the United States is not in the top 20 is “paying taxes,” where it ranks a miserable 72. (The U.S. ranking has shifted from 76 in the 2008 report to 46 in 2009 to 61 in 2010.) Tax reform that gets rid of the loopholes, deductions and credits — and the inherent corruption related to them — would clearly help the economy.

The point that taxes are low right now is well taken, as far as it goes. I think most small government advocates right now are not so much focused on lowering taxes as on reducing the deficit, about which Zakaria says nothing. The problem, of course, is that while taxes may be low now, everyone is afraid that they will rise. And that deters investment, entrepreneurship, risk taking, etc.

So, outside of the tax code, the United States does not seem to have slipped very much in terms of competitiveness and ease of doing business. What has changed? The answer is pretty clear. Only five years ago, American infrastructure used to be ranked in the top 10 by the World Economic Forum. Now we’re 24th. U.S. air infrastructure has gone from 12th to 31st, roads from eighth to 20th.

And so what? First, is this change because our infrastructure has gotten worse, or because other countries' has gotten better. But, more importantly, what is the evidence that infrastructure is an important constraint on the economy, at the margin? My own experience doesn't suggest that road quality is a significant constraint on the economy. The highways pretty much get me where I want to go. Potholes and other road quality issues are a negligible annoyance. If Zakaria wants to make this case, let alone to say that "the answer is pretty clear," he needs to cite data showing that companies are being significantly held back by the quality of infrastructure.

By the way, I think there might be some 21st-century infrastructure project that really would help to revive the economy-- something along the lines of universal free wi-fi, or better search engines, or a government-funded, much more comprehensive and smarter, Wikipedia, available through government kiosks anywhere, plus IP buyouts... But these guys always want to build the infrastructure for the last generation of technology, which we already have. It's a great plan... for the 1930s.

The drop in human capital is greater. The United States used to have the world’s largest percentage of college graduates. We’re now No. 14, according to the most recent OECD data, and American students routinely rank toward the bottom of the developed world. The situation in science education is more drastic. The number of engineering degrees conferred annually decreased more than 11 percent between 1989 and 2000. Even with the increase in college attendance over the past two decades, there were fewer engineering and engineering technologies graduates in 2009 (84,636) than in 1989 (85,002). Research and development spending has risen under Obama, but the basic trend has been downward for two decades. In percentage terms, the federal share of research spending — which funds basic science — is half of what it was in the 1950s.

Two problems here. First, decisions about how many people go into engineering and science are made by individuals. I suppose you could subsidize higher education in the natural sciences more, but that would just be helping people who are generally going to come out on top anyway. Second, it's up to the market to decide how much reward to provide to engineers, and thus, how much incentive to create for people to become engineers. My cousin got an engineering degree and it took him a year to get a job. He's doing all right now, but his experience does not suggest that demand is white-hot. It's just not smart to second-guess the market in such matters. If not that many people are going into engineering, the likely explanations are either that there's not really that much need for engineers after all, or that not many people have the aptitude and/or desire to do the job, even if the rewards are fairly good. Either way, the market outcome is optimal.

The irony is that Zakaria's own article is a good example of the ineptness in the use of numbers which he deplores.

In other words, the big shift in the United States over the past two decades is not a rise in regulations and taxation but a decline in investment — in physical and human capital. And investment is the crucial locomotive of long-term growth. Michael Spence, the Nobel Prize-winning economist, points out that the United States got out of the Great Depression because of the spending associated with World War II but also because during the war, it dramatically reduced its consumption and expanded investments. People spent less, saved more and bought war bonds. That surge in investment — by people and government — produced a generation of growth after the war. If we want the next generation of growth, we need a similarly serious strategy of investment.

I'm all in favor of investment! But the way to get it is to reduce the deficit so that investors are less afraid of future taxation and savings aren't being soaked up by government bonds (plus some loose money a la Scott Sumner). If Zakaria understood the ramifications of his own argument here, he'd be making common cause with those he mocks at the beginning. It makes me wonder: why is he taking this side of the argument at all? The numbers certainly don't compel him to it. I think the reason is his temperament and his social milieu: he likes elites, he likes arguments that empower elites, he likes arguments that ingratiate him to elites.

Anyway, back to where I started, my irritation. Zakaria's Olympian tone in "the data do not support it" is, to say the least, unwarranted. Maybe I should say insufferable.

Comments

[Nato]

The Economist says things like this all the time. This could come straight from a Lexington article.

[Nathan Smith]

Possibly, but I think their standards are usually a bit higher than this.

[Nathan Smith]

The Economist tends to be supercilious, but not quite as supercilious as Zakaria is here.

[nato]

"Second, it's up to the market to decide how much reward to provide to engineers, and thus, how much incentive to create for people to become engineers. My cousin got an engineering degree and it took him a year to get a job. He's doing all right now, but his experience does not suggest that demand is white-hot. It's just not smart to second-guess the market in such matters. If not that many people are going into engineering, the likely explanations are either that there's not really that much need for engineers after all, or that not many people have the aptitude and/or desire to do the job, even if the rewards are fairly good. Either way, the market outcome is optimal."

Here's an alternate story: Science and engineering are in very high demand, but it requires good schooling from a fairly young age to yield the best abilities in those who are less predisposed toward those fields. I'm frankly surprised that an engineer could go a year without getting a job, because we always struggle to find engineers and frequently hire kids right out of college. Other companies do this even more than we do. The scientific firms in our area are the same; struggling to find people.

Granted, if you're not willing to move to a few areas in the country in which these sorts of jobs cluster, it could be harder, but there's a huge mismatch right now.

If there was some easier ways for companies to pay for young people to get degrees in return for coming to work for them that might help, and some companies try to do that, but generally speaking I don't think there's an efficient market there.

[Nathan Smith]

Interesting, but one question: If engineers are so hard to find, why don't you just pay more? I think everyone knows that engineering is a somewhat well-paid career in which one has better chance of employment than, say, anthropology. They just don't like it very much. But if they knew they could walk out of college into six-figure jobs, I think a lot more people would do it. You might even lure some people who do have the aptitude but chose some other major to go back to school for a career switch. If it wouldn't be worth it to pay that much more, well, maybe the marginal product of an engineer is not enough to offset the disutility of being an engineer for more people.

Anyway, this doesn't get Zakaria off the hook. Wanting to promote more science and engineering in schools is not at all inconsistent with overregulation and deficits being a drag on the economy. That was gratuitous. There might even be a link between a shortage of engineers (if there is one, in the appropriate sense) and overregulation and deficits. Suppose firms did start offering recent college grads in engineering $200K / year. They'll be nearly in the top 1%, and "more taxes on the rich" might convince them to take the fun route and be an English major. Also, how much of engineers' time is spent thinking about how to satisfy environmental requirements?

Finally, it's interesting that "we always struggle to find engineers and frequently hire kids right out of college" is given as evidence of a "huge mismatch right now." Shouldn't that be normal? Isn't that what college is for?

[nato]

I think the problem is that the response time to raised prices is very slow relative to the individual firm's hiring cycle. That is, we can hire engineers away from other firms, but we can't so easily convince people to become engineers. In any case, the salaries paid to engineers is already pretty high. If anything, the reason prices aren't higher is because companies can't easily tell how good an engineer will be ahead of time, and lots of engineers aren't very good*. Once a good engineer is identified, they can assume they will make six figures before very long. Apple's star engineers make high six figures, for example.

I only have so much visibility on how other engineers experience regulations, but I would say the #1 government-related thing engineers find irritating is worrying about IP. Technicians at the like can typically take care of regulatory hassles. Perhaps it's different for mechanical engineers in heavy industry, though the only one I know well works for a copper mine in Utah and has never mentioned anything except safety regulations.

"...it's interesting that "we always struggle to find engineers and frequently hire kids right out of college" is given as evidence of a "huge mismatch right now." Shouldn't that be normal? Isn't that what college is for?"

Amongst engineers with relevant experience, it typically takes 4-6 months before they know enough to be productive. Amongst college grads, it will take 4-6 months before we have an idea if they're ever going to be productive. It's a huge risk that usually isn't worth the salary savings. Basically every role is critical, and one failure can set back the whole project. Lots of other firms in the Valley have more pure/greenfield development work that requires fewer special skills and can tolerate more uneven quality, but we can't. So the fact that we hire out of school is really an expression of the reality that we just have to take risks in order to have a shot at filling all the slots. Some projects get created largely to absorb and occupy junior engineers so we can use them later, and if the product doesn't go anywhere it's not a huge loss.

*Social awkwardness, poor language skills, prima-donna attitudes, and unwillingness to work within project priorities being problems that are far more common than lack of aptitude.

[Nathan Smith]

By the way, what kind of engineers are we talking about here? Software engineers?

[nato]

Electrical engineers, software engineers, and applied scientists.

[Nathan Smith]

Fascinating. In spite of liking my job quite a bit, I'm often tempted to drop it all and try to be a software engineer instead. I find that fascinating. Also, I wonder if I could steer some of those mostly-for-training-purposes projects my way. I could maybe find projects at my university for young software engineers to hone their skills on.

[nato]

I think the single biggest motivator in engineering is how much creative control one gets. Money tends to be somewhat secondary, though they go together.

Historical trends in complexity, however, has been slowly eroding how much creative control any one engineer can have. I do wonder if that has dissuaded some who would have become engineers.

[Nathan Smith]

Fascinating. That makes me wonder whether engineering firms should imitate universities. Academia tends to offer great scope for creativity. It does so by coupling teaching, which isn't all that creative since one mostly has to teach the same old things, with research and scholarship, which doesn't contribute to the university's revenues except maybe indirectly through affecting the school's reputation, but does make the job a lot more attractive. What if engineering firms established a norm of six hours per day working, three hours personal/intellectual development? Would that attract good people?

[nato]

Google does something very similar to that called "innovation time off", and it has worked very well for them. Google doesn't pay all that much more than other firms - sometimes less - but it does get a lot of the best engineers that way. The difficulty is in maintaining the high management discipline necessary to keep deadlines and so on from squeezing that sort of thing out. There's always a temptation to get the project done faster even though in the long term it increases the chances that someone will be able to lure away your engineers. Apple sells its engineers on the idea that they are going to make the best whatever in the world, and that keeps them working long hours.

[arizona gynecologist]

The principle here is that blaming your partner is the same as self-blame.

[Indian domain names]

Thank you for this good collection of posts..