Friday, November 30, 2007

Dangerously Innumerate

Scott McLeod spews some jibber-jabber:

However, some states seem to be more GDP-efficient than others. For example, Connecticut is ranked 29th in overall population and 23rd in overall GDP, but is the 4th-ranked state when it comes to GDP per capita (column G). In contrast, Alabama is ranked 23rd in overall population and 25th in overall GDP, but is the 45th-ranked state in terms of GDP per capita. Connecticut’s GDP over/under (column H) is +19 (23 minus 4). Alabama’s is -20 (25 minus 45). Connecticut appears to be a GDP overachiever, while Alabama seems to be an underachiever. Dollar for dollar, person for person, Connecticuters are contributing more to the overall national economy than Alabamans.

Huh? If you want to know how "efficient" a state's economy is, in the gross terms we're dealing with here, you can just look at the per capita number. All this subtracting one rank from the other is completely meaningless.

What's even more embarrassing is Ewan McIntosh and Doug Johnson chiming in to praise it.

Also, the states that have the highest per capita GSP is because they've either got a big financial sector or lots of extractive industries. Does banking require 21st century skills? I know mining doesn't.


Robert Jones said...

Oops! Well spotted Tom. Red faces all 'round.

Scott McLeod said...

Hey Tom! Good to hear from you!

I'll be the first to admit that I'm no economist (and, maybe, in this instance, no mathematician!). I concur that GDP per capita itself is a great efficiency measure for a state's economy. But I was looking for the over/under idea: the difference between a state's overall GDP rank and its GDP per capita rank can be stark. I don't know what the correct term would be for this contrast: I used 'over/under' - I'm open to other suggestions. Got any for me? If so, I'll try to find data and run 'em...

There are lots of reasons why a state's 'over/under' is low/high: we can probably list a few that are unique for each state.

I like the title of your post: very clever!

Robert Jones said...

OK, so I've been thinking about this some more.

I'm afraid Tom is right Scott - your post makes no sense. Why should there be any correlation at all between a state's GDP (which as you say is closely correlated to population) and it's GDP per capita? There are small rich states and big poor states, and all other combinations I guess. There is no reason to expect a correlation, so it makes no sense to try to draw some meaning from the difference between the two.

Scott McLeod said...

I did calculate correlations between overall GDP and GDP per capita in my spreadsheet (and mentioned them in my post). But the over/under number is not a correlation. Instead, it's just a simple subtraction between the two ranks.

I think the over/under idea is important in some way, shape, or form, rather than just relying on GDP per capita alone. In the video, for example, the idea of overall GDP is stressed again and again: "the U.S. is #1 so why worry?" seems to be the predominant tone of the video. But that mindset masks the internal inefficiencies that are in many of our states. Sure, Michigan may have a large overall GDP, but it sure isn't happy with how its state economy is going right now. That's where the GDP per capita idea comes in and where the issue of "what do we do to raise our GDP per capita?" arises. The second number illuminates the first. Together they have more power (in my mind, at least), than either does alone.

I'm not alarmist about the changes we're seeing in the global workforce, but neither do I feel that the U.S. can afford to sit on its heels complacently patting itself on the back. I think there's value in some kind of indicator that CONTRASTS a state's two GDP numbers, because neither alone shows the discrepancies between the two. I'm the first to admit I'm not an economist, so maybe the simple over/under number is not the right number to use. As I have said repeatedly, I'm open to other suggestions on how to show the DIFFERENCES between these two ideas of overall output and efficiency that are present in many states. Thanks for sharing your thoughts and challenging my thinking on this.

Robert Jones said...

Hi Scott. I'm glad that you're being so good-humoured about this. There's certainly nothing personal about my agreeing with Tom's assessment of your post. My involvement in this discussions is purely as a mathematician with a fair grounding in statistics. I couldn't possibly comment on the complexities of US economics.

You're saying that "over/under" does not refer to a correlation, but in your post you said that some states have "overall GDP ranks and GDP per capita ranks that are congruent" whilst others have a "discrepancy". Surely that assumes that you are expecting some connection between them. Otherwise it's like saying "I find that my shoe size and the temperature are congruent today" on a chilly day, then saying "I find a discrepancy between my shoe size and the temperature" on a hot day.

Let me put it another way. On your spreadsheet, Alaska has an "over/under" of plus 42, whilst New York has minus 7. The simple reason for this is that New York has a big population, so its GDP rank is only 3, whereas Alaska has a small population, so its GDP rank is 45. They have vaguely similar GDP per capita, so they have very different "over/under"s. End of story. Do you think that we should read something else into the difference?

Scott McLeod said...

Great points. I think the issue may be whether we believe that a big state also can have a high GDP per capita (as you're saying, should we really EXPECT both overall GDP and GDP per capita to be high?). When I was thinking 'congruence,' I was thinking 'Sure, there are a lot of people in some of these states, but is their economy humming or bumming?' For many of the larger states, it's the latter (e.g., Michigan, Ohio).

I think it's possible for large states to have an economy that's also efficient (as measured by GDP per capita). New York is a good example (at -4). So are New Jersey and Virginia. Obviously everyone wants higher GDP per capita. If Michigan and Ohio could up theirs, their overall GDP would be even higher because they'd be getting more economic output per person.

I think the goal is a high GDP per capita AND a high overall GDP. Small states may always have an efficiency advantage over larger states, just like smaller countries may always have an efficiency advantage over larger countries, but it is possible to have both. And that's really the point that I was trying to get at, which was what's the best remedy for a state with an ailing economy (GDP per capita) despite a high overall GDP? In my mind, it's strategic investments in digital technologies and 21st century skills in K-12 education.