« Status competition, rural Indian style | Main | The best sentence I read today »
Assorted
1. On inequality, Krugman responds to critics and Samwick adds further commentary. I'll note that the "marginal products" of big changes in government, society, technology, etc. are not always well-defined.
2. Tower Records is bankrupt again, and this time the stores may not survive.
3. In case you missed it, there is now very strong evidence for the existence of "dark matter."
4. The genetic causes of autism -- do they lead to early brain inflammation? Have I mentioned that my mother was instrumental in founding and running a care home for autistic children? Among other things, I use this blog to send her the latest news on the topic.
5. Seven puzzles: find them here, with solutions, and one of them is explained by GeekPress.
6. Virginia Postrel in Forbes, on why median incomes are not stagnating, here is a summary and a link.
Posted by Tyler Cowen on August 22, 2006 at 01:49 PM in Web/Tech | Permalink
Comments
Interesting links.
(Spoiler warning on link 5 #3)
I am confused about the solution to the dot-town suicides. It seems that his solution is incorrect, maybe someone can explain to me why it isn't. Here is a counter example: In a population of 100, there are 33 blues and 67 reds. The stranger comes in and says "I see that the number of blue dots is not a multiple of 6". This is nontrivial. So the people with blue dots look around, they see 32 blue dots. The people with red dots look around, they see 33 blue dots. No one is expecting there to be a possiblity of anyone killing themselves the next night. The lack of deaths does NOT expand the set of "forbidden" numbers because the lack of deaths does not convey any extra information to anyone.
Posted by: Agent00yak at Aug 22, 2006 3:21:30 PM
Agent00yak:
Keep in mind that every citizen knows that there are both red dots and blue dots in the city. Now, let's change the initial split from 33b/67r to 31b/69r. The strangers says "I see that the number of blue dots is not a multiple of 6." On that first night every blue looks around and sees 30 blue dots (which is a multiple of 6) and thus concludes that he must have a blue dot on his own head (and thus commits suicide). OK so far?
OK, now let's go to 32b/68r. On the first night no one kills themselves right? BUT the citizens can then be sure that the split is not 31/69 because if that was the split all the blues would have killed themselves the first night. It must be a different split, and the split must be 32/68. But the blues only see 31 blue dots, so they know they are the 32nd! So they must kill themselves!
But wait, not lets make it 33/67 (as you said). You can get from 32/68 to 33/67 in the exact same way. Everytime the group doesn't commit mass suicide it eliminates a possible split of red/blue until only one possible split remains and then the mass suicide happens.
Posted by: harryh at Aug 22, 2006 4:16:59 PM
Incidentally, a very interesting thing that the stranger can say:
"There are both blue dots and red dots in your city!"
Everyone already knew this! But it will still lead to mass suicide!
Posted by: harryh at Aug 22, 2006 4:22:22 PM
Krugman's response - there is no data to support blaming IT for wage inequality is rather weak considering that there is no evidence that politics is to blame for income inequality.
Its also odd that the administrations that he praises FDR, Carter and the like were not exactly booming times for the US economy.
Additionally, I have yet to hear a single explanation on why income inequality matters. Is it because we are all doing better so this is as good as a complaint as they can find?
Posted by: Chris at Aug 22, 2006 4:41:46 PM
Chris:
considering that there is no evidence that politics is to blame for income inequality.
Doesn't Krugman present several pieces of evidence in the e-mail MR links to? It would certainly be reasonable to refute this evidence as incorrect, but to say he offeres no evidence is incorrect.
Additionally, I have yet to hear a single explanation on why income inequality matters.
I have seen quite a few studies that attempt to show that people's happiness is more closely linked with how well off they are relative to their peers, than how well off they are on some sort of absolute scale. This sort of research rings true to me. I don't miss my flying car, but I might if all my neighbors had one (even if I was driving around on the ground in a Mercedes).
Posted by: harryh at Aug 22, 2006 4:52:15 PM
Chris - depending on which data you look at and how you interpret it - we are not all doing better. Even if you take the best case, faulty cpi calculations, fringe benefits, family size changes, etc, the increase for low to median households over the last 30 years is very small. In fact if you look at gains in average income from the IRS (averaged of tax returns)see
http://www.visualizingeconomics.com/wp-content/uploads/avg_income_book_big.gif
our performance is very poor since the early 70's. The GDP real growth per working person has been growing about 1.7% per year since 1980. but this is calculated using the GDP deflator not cpi. If you have educational or medical cost you are worse off that cpi indicates.
Posted by: joan at Aug 22, 2006 5:46:26 PM
Krugman's response has a bit of merit, but not much.
He's certainly right that, technology being exogonous to most economic models, it tends to become a catch-all for any effects that can't be explained otherwise. The direct evidence that technological change has fed income inequality is indeed weak. On the other had, the "evidence" Krugman cites that government policy is to blame is pretty weak, too.
He notes that growth in inequality is correlated with Republican presidents. Even he calls this "mysterious", which as far as I can tell, means "this doesn't take into account lag times, the contributes of congress, or any specific policies, so I can't begin to defend it with any intellectual rigor, but it does fit well with my gut feelings."
He notes that growth in inequality is large in the U.S., less in the U.K., and even less in other first world countries. Of course, that's the ranking of overall growth, too, so that fact could equally well be used to support the idea that technology-driven productivity changes are to blame.
As for specific mechanisms, he mentions unions and corporate taxes. But union membership has been plumeting steadily since the 60s; it's hard to point to any specific goverment policies that have driven that. And even if all the money freed up by reduced corporate taxes flowed into stockholder's pockets, that's not enough to explain the tremendous gains in the income of the top 1% and 0.1%.
Finally, Krugman makes some noises about social norms that favored equality in the post-war period and a reduced "climate of scrutiny" today. This might even be true, but it is speculation without even a hint of data and it isn't even about government policy.
Posted by: David Wright at Aug 22, 2006 8:00:29 PM
Actually Cris, under FDR real per capita gdp growth averaged some
9%-- compared to a long term average of under 2%
Under FDR real growth was at least twice as strong and actually more like 4 to 5 times as strong as under any other
president reguardless of what measure you use.
What other great insights do you have for us?
Posted by: spencer at Aug 22, 2006 8:28:25 PM
"Among other things, I use this blog to send her the latest news on the topic."
Heh, I always suspected your apparently fascination with autism-spectrum disorders was for personal reasons but that wasn't the one I would have guessed... :) :) :)
Posted by: Jacqueline at Aug 23, 2006 6:01:17 AM
David Wright writes:
But union membership has been plumeting steadily since the 60s; it's hard to point to any specific goverment policies that have driven that.
Government constraints on union organizing. In the US, unlike many other industrialized nations, workers eeking union representation have to go through a multi-step procedure that can drag on for years. (By way of comparison, if people who wanted to form a Delaware corporation had to meet for a week in Wilmington, there would be a lot fewer Delaware corporations.) A series of decisions unfavorable to union organizing by the Republican-dominated NLRB over the past 25 years has made the process even more difficult.
(If you think unions are bad, you may think that's a good idea, but the point is that there is an identifiable government policy at work here.)
Posted by: alkali at Aug 23, 2006 8:45:07 AM
"Actually Cris, under FDR real per capita gdp growth averaged some
9%-- compared to a long term average of under 2%
Under FDR real growth was at least twice as strong and actually more like 4 to 5 times as strong as under any other
president reguardless of what measure you use.
What other great insights do you have for us?"
Amazing what joining a word war will do to right after a depression.
Posted by: Tom at Aug 23, 2006 10:51:53 AM
Sorry. WORLD war
Posted by: Tom at Aug 23, 2006 10:53:06 AM
Wartime prosperity reconsidered
Posted by: TGGP at Aug 23, 2006 1:29:00 PM
harryh: It seems there is an assumption in there that people of Dot Town a: Wouldn't possibly kill themselves during town gathering. and therefore b: People all think things through at the same speed. Without these two assumptions it doesn't seem like the data could propagate through the population.
Posted by: agent00yak at Aug 23, 2006 1:41:10 PM
David Wright--"He notes that growth in inequality is large in the U.S., less in the U.K., and even less in other first world countries. Of course, that's the ranking of overall growth, too, so that fact could equally well be used to support the idea that technology-driven productivity changes are to blame."
Productivity is measured by GDP per worker. On this measure the US is about average for first world countries since 1980. Norway leads the list with 2.4% annual growth in GDP/worker the US is at 1.7%. Our larger GDP growth is due to greater population growth. (see)
ftp://ftp.bls.gov/pub/special.requests/ForeignLabor/flsgdp.txt
Posted by: joan at Aug 23, 2006 1:54:29 PM
Joan: I am not confused by per capita vs. aggregate. Here are the figures for growth in real GDP per capita from 1980-2000 for 5 of the major industrial economies: U.S. 104%, U.K. 55%, Germany 43%, Italy 43%, France 38%. As you can see, in the time period you chose, the figures behave exactly as I characterized them. Yes, you can find a few of the smaller European countries -- I would have picked Luxemburg at 146% over Norway with just 61% -- that have U.S. levels of growth, but I think that obscures the broad picture of how European policies have affected the standard of living for the vast majority of Europeans.
Posted by: David Wright at Aug 23, 2006 3:36:31 PM
Joan: To follow up, I did the computation again with the BLS figures you cited. Apparently you prefer GDP/worker to GDP/capita, whereas I regard that metric as concealing the fact that the Europeans pay their least productive workers to stay at home. In any case, even your prefered metric supports my claim, although admitedly less strongly than my prefered metric. Of the big countries, the U.S. and the U.K. are on top with 55% and 60% growth in GDP/worker from 1980-2005, followed by France with 50% growth, Italy with 35% growth, and Germany with 25% growth. Yes, Norway is at 70%, but you'll notice that growth comes predominately in the last 5 years as the oil price surge has brought them a tremendous windfall; that's hardly an argument for Norwegian labor market policy. (Luxemburg, I see, isn't on your list.)
Posted by: David Wright at Aug 23, 2006 4:47:22 PM
Hi all! This a good site! Cheap Buy Abilify online order http://galeon.com/abilify/ abilify
Posted by: Abilify at Aug 23, 2006 5:33:43 PM
US real GDP/capita in 2002 dollars
1980 23615
2000 36215
2005 39103
How do you get 104%
Posted by: joan at Aug 23, 2006 5:45:12 PM
Joan: Sorry, that should have been 56%. The figures for the other countries are correct. I admit that mistake makes me look bad, but the overall pattern does still hold. (Here is where I'm getting my data. I orginally did the computations starting in 1970, but then changed to 1980 when I saw that was your prefered start date. I then neglected to change the initial value for the U.S.) Thanks for the careful reading.
Posted by: David Wright at Aug 23, 2006 6:01:36 PM
Hi all! This a good site! http://galeon.com/abilify/ Abilify [url=http://galeon.com/abilify/]Abilify[/url]
Posted by: abilify at Aug 26, 2006 6:52:45 AM
Posted by: at Oct 14, 2008 1:57:43 AM
Replica Gucci Jewelry
Gucci Replica Jewelry
Tiffany Jewelry
Replica Tiffany Jewelry
Posted by: aion kina at Mar 20, 2009 9:26:30 PM
Mabinogi online gold
Mabinogi money
Mabinogi Gold
cheap Mabinogi gold
buy Mabinogi gold
2moons dil
2moons gold
buy 2moons dil
2moon dil
cheap 2moons dil
Flyff gold
flyff Penya
flyff money
buy flyff penya
cheap flyff penya
cheap flyff gold
aion gold
buy aion gold
cheap aion money
aion money
cheap aion gold
Dofus kamas
buy dofus kamas
cheap kamas
dofus kama
dofus gold
dofus money
Knight Online Gold
Knight Gold
Knight Noah
Knight Online Noah
Posted by: aion at Jul 9, 2009 1:41:17 AM
