« How to learn about everything? | Main | Betting markets in everything »
Ask Marilyn: IQ vs. the economists
Marilyn vos Savant has (supposedly, read the link) the world's highest recorded IQ at 228. She is now fielding questions on economics. Here is one example:
Question: Why has the income disparity grown so much in developed countries? – Matthew Cencich, Victoria, British Columbia, Canada
Answer: I think the disparity is a normal result of overall economic growth. The bottom incomes (zero) can’t go lower, but the top incomes can go up and up. And so they do, of course.
Here is another:
Question: Do you think that government actively encouraging people to borrow money (and spend it) is the right way to resolve the recession? – Caroline Kelly, Hendon, London
Answer: No. I think that more consumer activity could be modestly helpful to the economy in the short term. And in better times, it would even masquerade as growth. But in the current climate, increasing the family debt would cause added personal financial problems. So I doubt that it’s useful for government to advocate shopping as a form of national service unless elected officials are perhaps a tad more interested in shifting a little of the burden away from themselves than they are in a lasting solution.
She also explains the financial crisis. ("First, an economy based on growth is bound to falter now and then. A whole sector could collapse. It’s a house of cards. Eventually, it must morph into a system that functions on stability, or it will fail – meaning a fall large enough to cause an unstoppable breakdown and widespread hardship.")
Question: on these problems, does she do better or worse than leading economists?
Addendum: Here is Marilyn on YouTube. Oddly (look just past the 4:00 mark), she can't define IQ properly.
Posted by Tyler Cowen on May 31, 2009 at 06:27 AM in Education | Permalink
Comments
I was with her until I followed the youtube link and found her trotting out the whole "we only use 10% of our brains" trope.
Posted by: Trent McBride at May 31, 2009 8:36:53 AM
She doesn't say anything stupid, fallacious, or transparently self-serving, so I'll take it.
Posted by: Keith at May 31, 2009 9:39:53 AM
I've read her before, and she is clearly an intelligent and widely read person. However, I would appreciate it if she learned when to just say, "I don't know." She completely misses in your first example:
Q: "...income disparity grown so much in developed countries?"
A: "...The bottom incomes (zero) can’t go lower, but the top incomes can go up and up."
Incomes of zero are not included in (any reasonable) income disparity measures.
Posted by: Steve at May 31, 2009 10:15:47 AM
@Steve: whatever epsilon you count as the bottom, it rounds to zero.
Posted by: Phil at May 31, 2009 10:19:22 AM
She also says "We only use 5% of our brains." I was told that's a myth.
https://healthlibrary.epnet.com/GetContent.aspx?token=c5987b1e-add7-403a-b817-b3efe6109265&chunkiid=156966
Was that the accepted scientific belief at the time?
Posted by: david at May 31, 2009 10:24:11 AM
The first question is based on a false premise. There is no systematic relationship between national income and income disparity. The gini coefficients of feudal economies is way higher than a modern developed one. The proper answer to such a question is "mu".
Also, my understanding is that wealth is much more unevenly distributed than income. To the more specific question of why wealth disparities have grown in the United States after say 1960, her answer contributes nothing, because few Americans were living at the subsistence level.
Posted by: OneEyedMan at May 31, 2009 10:25:22 AM
With regard to increasing wage inequality--why has its increase been more dramatic in the US than Japan? Indeed, I have seen one study that showed that the modest increase in wage inequality in Japan (relative to the US) can be traced to the increasing age of the Japanese workforce. The answer to this question is taboo.
Posted by: Richard A. at May 31, 2009 10:34:49 AM
Who can define measurements of intelligence properly?
Posted by: Diversity at May 31, 2009 10:41:10 AM
Is the complaint here that she disagrees with economic theory, or that she doesn't understand economic theory? Those are very different things. (I don't think we can distinguish between them with the current evidence.) Also, she is simplifying things for an unsophisticated audience, which adds a complication.
Posted by: o at May 31, 2009 12:38:30 PM
While I find Ms Vos Savant's observation of the set limits of income quite clear, I don't believe that said set limits are causative. They merely limit the possibilities. It is harder for "nothing" to be smaller then it is for "something" to get bigger.
The comment that is more likely close to being explanatory is offered by Richard A as a question. "Why in the US, but not elsewhere?".
Posted by: Hiero2 at May 31, 2009 12:44:34 PM
As to whose comments are more intelligent - there is no question - vos Savant wins by definition!
As for whether Ms vos Savant delivers better commentary than economists, I would rate her comment on the financial crisis as about as rational as the Austrians, but rather less than economic commentary available elsewhere (like here).
However, as we see so clearly in hindsight, neither intelligence nor education carry crystal balls. Personally, I think our economists and our financiers, as two collective groups, dropped the ball big time on this one. I do not count myself as any better, but at least it wasn't in my job description.
I'd like to ask a question in return. Given that for the derivatives market, along with other contributory factors, risk level skyrocketed after Bush took office (isn't it really far too coincidental to assume that Greenspan and Freddie Mac took their actions in a vacuum, i.e. independently of the White House leadership?), how long would a collapse have been delayed if policy had been maintained approximately as it was in the Clinton years? How big could the collapse have been if it had occurred in 2001, the year after Bush took office?
Posted by: Hiero2 at May 31, 2009 1:05:29 PM
"Who can define measurements of intelligence properly?"
Shane Legg and Marcus Hutter do so formally here:
http://www.hutter1.net/ai/iors.pdf
My abstract definition: Take the set of all possible problems, where a problem includes an environment, set of resources and constraints. The subset of these problems that an agent can solve relative to the subset that the agent can not solve defines the agents intelligence.
If she wants the highest IQ claim, she should have a contest with:
http://en.wikipedia.org/wiki/Terence_Tao
Posted by: Rob Sperry at May 31, 2009 1:09:58 PM
Rob,
Problems with your approach;
1. A solves problems 1,3, and 8 out of 10. B solves 2, 5, and 6. Who is more intelligent?
2. Out of a thousand "test takers" the average number solved is 5, but 999/1000 cannot solve #9 and #10. One person can only solve 9 and 10. Who is more intelligent?
3. The test would have to include non-solvable problems. How many points do you get for recognizing this? How many points for persistance? Who decides that they are non-solvable
4. etc etc
Posted by: Robb Lutton at May 31, 2009 1:33:37 PM
"Personally, I think our economists and our financiers, as two collective groups, dropped the ball big time on this one. I do not count myself as any better, but at least it wasn't in my job description."
While some economists beat themselves up as a group, many having been talking about growing imbalances that must be unwound since the early 1990s. For some reason these "dropped the ball" criticisms have a whiff of kinship with criticisms of hard sciences for not knowing or fully understanding something before something was known or fully understood.
How big could the collapse have been if it had occurred in 2001, the year after Bush took office??
You're conflating younger sparks with older tinder.
Posted by: PM at May 31, 2009 2:46:50 PM
There are uncountably many problems (using some reasonable definition), but only countably many solutions (each solution can be expressed as a finite string in some finite alphabet). So almost every problem is unsolvable.
Posted by: Andy at May 31, 2009 3:00:18 PM
Hiero2 touches on a very serious issue. It is this desire by the public to have economists play GOD that I find most disturbing. It shows off base the economics profession is in properly communicating to the public the benefits of economic education.
Posted by: John Pertz at May 31, 2009 3:26:44 PM
Robb-
I agree with your objections. One way Legg and Hutter propose to measure the importance of problems is by their complexity. This is has the benefit of being theoretically tractable and objective. But there is also a sense in which we might want to order or weigh problems by value, which is implied in part 1 and 2 of your objection. I don't think there is going to be a math like solution for that. But for simplicity maybe we can restrict value weights to wisdom and leave it out of raw intelligence.
Posted by: Rob Sperry at May 31, 2009 3:39:30 PM
Anyone else notice that she clearly doesn't know what a parameter is?
Posted by: Robert Simmons at May 31, 2009 4:53:36 PM
Andy,
There are uncountably many problems (using some reasonable definition), but only countably many solutions (each solution can be expressed as a finite string in some finite alphabet). So almost every problem is unsolvable.
Does that argument still work if more than one problem can have the same solution?
Posted by: rf at May 31, 2009 5:53:10 PM
Marilyn vos Savant may have a very high IQ, but she also has a high Agreeableness and low Openness, which means she doesn't have anything interesting to say. And she's also kind of a failure in life because, as far as I know, she hasn't used her IQ to accomplish anything besides writing her silly column.
Posted by: Half Sigma at May 31, 2009 6:43:13 PM
Better then Krugman but I suspect Ms vos Savant has gaps in her economic readings
Posted by: DanC at May 31, 2009 10:43:06 PM
To the more specific question of why wealth disparities have grown in the United States after say 1960, her answer contributes nothing,
On the contrary. Her answer applies much more to wealth than income.
The USA really doesn't have a significant group of zero income earners. (outside of children and like) But a very significant proportion of the population has zero (or negative) net wealth.
Posted by: doctorpat at May 31, 2009 10:43:43 PM
Her column is syndicated in Parade magazine. Nuff Said.
Posted by: Newsie at Jun 1, 2009 7:39:34 AM
I recall reading a paper whose conclusion was that the wealth distribution was close to an exponential, the expected outcome if wealth was distributed randomly.
Posted by: Cyrus at Jun 1, 2009 8:28:45 AM
"But a very significant proportion of the population has zero (or negative) net wealth."
But wealth can go below zero. So disparities can be created as much by growth as the opposite.
Posted by: assman at Jun 1, 2009 9:33:06 AM
