Politics, Economics and other Stuff

Living in the Boondocks

Well, in my former New Haven suburb I get home delivery of the NYT, WSJ and FT.  My new very upper middle class Fairfax County suburb doesn't seem to have home delivery of the NYT or WSJ.  I've not tried getting the FT yet, but I don't think home delivery is likely if the NYT or WSJ doesn't have it. Very surprising, even weird. I could try work delivery, but that sort of undermines the point of getting a paper.  But since I don't have a train commute anymore print newspapers may be on the way out in any case. 

Late Payments and FICO

A 30 day late payment is almost as costly as foreclosure to you credit score (as known as FICO).   I had no idea --- I had wondered about this and I now have a professional reason to care so I might take a look...seems weird.    This would explain some of the extreme white-black credit score differences -- the underlying behavior may not be as different as I though it had to be.  I reminder never to be 30 days late (i.e. not paying with in a few weeks of getting a 2nd notice).  (Yahoo via Volokh) 

Poverty in the DC metro area

The WP writes:

One in four students in Fairfax County qualifies for free or reduced-price meals this fall, up from one in five three years ago. In Montgomery County, 29 percent of students were deemed eligible for meal subsidies in October, up from 26 percent in October 2007. In Prince William County, the eligibility rate increased from 29 to 33 percent, and in Prince George's County, from 46 to 52 percent. ...

Many parents who work as housecleaners, construction workers or landscapers have lost their jobs, or their hours have been cut.

It was my impression that most of the social strata that is poor or near-poor or poverty endangered doesn't live in Fairfax County and commutes in, but I should look more closely at the relevant survey data.  Also, who puts out the free or reduced-price meal school lunch data?  How is it updated?  This might be an interesting data source beyond unemployment to get at the geographical organization of financial distress. Ah, here is the data, not clear if the frequency or detail is useful as published.  The WP helpfully doesn't cite any data sources.  BTW, 185% of poverty level or lower gets you free or reduced-price meals. 

Not that high participation in food support programs is new -- what surprised me here is the high numbers for one of the highest income counties in the US. For instance, Andrew Gelman writes:

Mark Rank and Thomas Hirschl recently published an estimate that 50% of American kids are on food stamps at some point during their first twenty years of life. Their estimate is based on an analysis of data from the Panel Study of Income Dynamics, from 1968 through 1997.  ... 

This is all fine--it's an efficient use of the data they have--but I'd feel a bit more confidence in Rank and Hirschl's estimates if they would cross-check by doing some raw-data calculations based on the subset of kids who were in the study continuously for ages 1-20. That's a crucial component in any applied statistical analysis--the continuous thread connecting the raw numbers to the final estimate--and I always like to see it, especially for a politically-charged subject such as this one. ...

Here are the key results from the study:

(from Table 1): 12% of newborns were on food stamps. 49% of kids were on food stamps for at least one year between ages 1-20. 23% of kids were on food stamps for at least 5 years.

(from Table 2); 8% of white newborns and 33% of black newborns were on food stamps. 37% of white kids and 90% of black kids were on food stamps for at least one year between ages 1-20.

(from Table 3): Among the black kids of unmarried parents where the head of household did not graduate from high school, 99.6% were on food stamps.

Again, I don't know how much to believe these numbers, but I assume that they're not too far from what was really happening in those years.

Also, whassup with those superfluous decimal places? "22.8%" and all the rest? Doesn't anybody teach these people about sampling variation and significant digits?

The NYT writes about the surge in food stamp program participation.  Not sure if the NYT coverage is insightful as opposed to sensationalist, but recent program changes do seem to bear looking into in terms of program eligibility, program outreach and program uptake.   This is not a small program.  The NYT writes:

While use is greatest where poverty runs deep, the growth has been especially swift in once-prosperous places hit by the housing bust. There are about 50 small counties and a dozen sizable ones where the rolls have doubled in the last two years. In another 205 counties, they have risen by at least two-thirds. These places with soaring rolls include populous Riverside County, Calif., most of greater Phoenix and Las Vegas, a ring of affluent Atlanta suburbs, and a 150-mile stretch of southwest Florida from Bradenton to the Everglades.

Although the program is growing at a record rate, the federal official who oversees it would like it to grow even faster.

“I think the response of the program has been tremendous,” said Kevin Concannon, an under secretary of agriculture, “but we’re mindful that there are another 15, 16 million who could benefit.”

Nationwide, food stamps reach about two-thirds of those eligible, with rates ranging from an estimated 50 percent in California to 98 percent in Missouri. ...

Now nearly 12 percent of Americans receive aid — 28 percent of blacks, 15 percent of Latinos and 8 percent of whites. Benefits average about $130 a month for each person in the household, but vary with shelter and child care costs.

In the promotion of the program, critics see a sleight of hand.

“Some people like to camouflage this by calling it a nutrition program, but it’s really not different from cash welfare,” said Robert Rector of the Heritage Foundation, whose views have a following among conservatives on Capitol Hill. “Food stamps is quasi money.”

Official summary program data here. 

Nativism

Razib writes:

One of the notable features of North American history and culture is that it is a mosaic of different populations, and, that mosaic has come about in very different ways. For example, the millions of Italian and Jewish Americans descend from hundreds of thousands of Italian and Jewish immigrants. By contrast, millions of Yankees and Quebecois descend from tens of thousands of ancestors, who arrived in the 17th and 18th centuries. In 1992 a Census demographer, Gibson Campbell, calculated that 49% of the population of the United States in 1990 was descended from those whose ancestors were resident within the United States in in 1790 in "The Contribution of Immigration to the Growth and Ethnic Diversity of the American Population" (inclusive of blacks and whites). 51% were descended from those who arrived after 1790. Put it another way, 127 million Americans in 1990 were attributable to the net 50 million immigrants who arrived after 1790. The remainder of the population would be attributable to the 4 million U.S. residents in 1790.  [Net Italian immigration is above a million.]

Since I think of my self as half American and half European, I just noticed that my kids are by decent 3/4th pre-1700 immigrant to North America.  At least as far as my awareness of the genealogy is correct. 

Update #1: Jody comments:

All my ancestors are post-1790, while about half of Spouse's ancestors are pre-1790. (That percentage may be higher -- spouse's family tree fades into the unknowable past on many branches at approximately 1830. The known branches march back to the Mayflower, to Rhode Island in the 1640s, and to Virginia in the 1750s.)

This reminds me: how do they calculate the percentages? With intermarriage between colonial settlers and later immigrant groups on the rise throughout the 20th century, does it only take one ancestor to put you on the other side of that 1790 divide? How would Spouse have been counted, given that his family name comes from Mormon immigrants who arrived in 1861, but his second great grandfather married into an old New York family who traced their arrival to the Mayflower? How would our children be counted? It makes sense that the early arrivals would be over-weighted, if any percentage counts.

I tried finding a public version of the paper since I no longer have access to JSTOR -- I'm pretty sure the fraction is by (estimated) genetic contribution, otherwise the numbers seem (far) too low for people with some pre-1790 ancestery.  That's how I'm calculating the fraction for my kids.  The numbers may have changed for more recent genetic analysis of this question.  

Oregon in the NYT

Camps of homeless runaway kids this time.  Despite the headline 'With More Troubled Families, More Runaways' it's not surprising that there is no actual data:

Federal studies and experts in the field have estimated that at least 1.6 million juveniles run away or are thrown out of their homes annually. But most of those return home within a week, and the government does not conduct a comprehensive or current count.

The best measure of the problem may be the number of contacts with runaways that federally-financed outreach programs make, which rose to 761,000 in 2008 from 550,000 in 2002, when current methods of counting began. (The number fell in 2007, but rose sharply again last year, and the number of federal outreach programs has been fairly steady throughout the period.)

'Advanced' Math Abilities

The just released national 4th grade math test results are, as usual, scary.  These national standards -- and I'm not going to get into the measurement issues here -- have four ability categories: below basic, basic, proficient, advanced. 

8% of whites, 1% of blacks and 17% of Asians score as advanced.  These are huge differentials.  Asians outperform blacks 1600%, not something like 80%.  Even whites outperform blacks by a factor of 8.  Free school lunch eligible children are 1% advanced, non-school lunch eligible children are 10% advanced in math. Again, a frightful situation given the number of free school lunch eligible children, at least 18 million at any one time, 30 million over the full school year.  Not clear what fraction of all enrolled students that is, but I'd hazard around 1/3rd. A huge fraction of kids is just being cut out of opportunities here.

Update #1: 'school lunch eligible' may mean reduced price eligible, not eligible for free lunch.  Which would explain the large numbers of kids.  Still, not over 200% of the poverty line either, I think.  Hard to find the demographics online, at least I don't know where to look. 

Don't get carried away with those high Roman population estimates

Not that this is necessarily news.

Razib writes:

Peter Turchin is at it again, Coin hoards speak of population declines in Ancient Rome (ungated version):

In times of violence, people tend to hide their valuables, which are later recovered unless the owners had been killed or driven away. Thus, the temporal distribution of unrecovered coin hoards is an excellent proxy for the intensity of internal warfare. We use this relationship to resolve a long-standing controversy in Roman history. Depending on who was counted in the early Imperial censuses (adult males or the entire citizenry including women and minors), the Roman citizen population of Italy either declined, or more than doubled, during the first century BCE. This period was characterized by a series of civil wars, and historical evidence indicates that high levels of sociopolitical instability are associated with demographic contractions. We fitted a simple model quantifying the effect of instability (proxied by hoard frequency) on population dynamics to the data before 100 BCE. The model predicts declining population after 100 BCE. This suggests that the vigorous growth scenario is highly implausible.

The figure to the left shows the reasoning. A simple model which related population size (dependent) to coin hordes (independent) was fitted before 100 BCE. The correlation between coin hordes to population size and political stability are well attested for many polities. In any case, using the model and projecting outward with the coin hordes known for the early imperial period a theory which suggests that multiplicative increases in census size during the Julio-Claudian age were a function of a shift in the accounting method (instead of simply males, including the whole household) was supported. The high count was already implausible on other grounds (e.g., if true, that means that Italy never attained the early Roman imperial population size until the mid-19th century), but that the model fits so well with the lower projections previously offered by other scholars is very suggestive. Contra extreme subjectivists some models of the past are probably right, and some are probably wrong.

The Ancestors of African Americans

To follow up on the post below, I tried to google recent articles on African American genetic origins. 

Lind et. al., Elevated male European and female African contributions to the genomes of African American individuals, Human Genetics (2007).

The differential relative contribution of males and females from Africa and Europe to individual African American genomes is relevant to mapping genes utilizing admixture analysis. The assessment of ancestral population contributions to the four types of genomic DNA (autosomes, X and Y chromosomes, and mitochondrial) with their differing modes of inheritance is most easily addressed in males. A thorough evaluation of 93 African American males for 2,018 autosomal single nucleotide polymorphic (SNP) markers, 121 X chromosome SNPs, 10 Y chromosome haplogroups specified by SNPs, and six haplogroup defining mtDNA SNPs is presented. A distinct lack of correlation observed between the X chromosome and the autosomal admixture fractions supports separate treatment of these chromosomes in admixture-based gene mapping applications. The European genetic contributions were highest (and African lowest) for the Y chromosome (28.46%), followed by the autosomes (19.99%), then the X chromosome (12.11%), and the mtDNA (8.51%). The relative order of admixture fractions in the genomic compartments validates previous studies that suggested sex-biased gene flow with elevated European male and African female contributions. There is a threefold higher European male contribution compared with European females (Y chromosome vs. mtDNA) to the genomes of African American individuals meaning that admixture-based gene discovery will have the most power for the autosomes and will be more limited for X chromosome analysis.

I'm a bit doubtful about fractions to 1/10,000 with a sample of 93...still, supports the ballpark figure of a bit under 1/3rd for European ancestry for male DNA.  1/5th for autosomal DNA.    

Tishkoff et.al., The Genetic Structure and History of Africans and African Americans, Science (2009)

Africa is the source of all modern humans, but characterization of genetic variation and of relationships among populations across the continent has been enigmatic. We studied 121 African populations, four African American populations, and 60 non-African populations for patterns of variation at 1327 nuclear microsatellite and insertion/deletion markers. We identified 14 ancestral population clusters in Africa that correlate with self-described ethnicity and shared cultural and/or linguistic properties. We observed high levels of mixed ancestry in most populations, reflecting historical migration events across the continent. Our data also provide evidence for shared ancestry among geographically diverse hunter-gatherer populations (Khoesan speakers and Pygmies). The ancestry of African Americans is predominantly from Niger-Kordofanian (~71%), European (~13%), and other African (~8%) populations, although admixture levels varied considerably among individuals. This study helps tease apart the complex evolutionary history of Africans and African Americans, aiding both anthropological and genetic epidemiologic studies.

So a lower figure, 13% European, 8% other non-African.  Indeed, it appears that recent studies have lead to downward revisions of the estimated European ancestry of African Americans from what they were a even decade ago. Tishkoff et.al. also find very low (essentially no?) Native American ancestry in their sample of African Americans.  Which raises the issue of getting a good sample of African Americans, if there's lots of heterogeneity. 

I'm also revising my estimate of how much African DNA is US self-identified whites down...there's no presumption that there's symmetry in gene flow.  That's not how this works.  So were' probably under 10 million but over a million (weighting by fraction -- we're in the tens of millions if you look for whites with small amounts of African ancestry). 

Where people without health insurance live

Floyd Norris of the NYT writes (via. Razib):

Those who lack health insurance now are far more likely to live in states that usually vote Republican — the states whose senators and representatives are least likely to support a law to extend coverage.

That would seem to indicate that Republican constituents are the ones who would most benefit from passage of universal health insurance coverage. But an analysis of Congressional districts within those states indicates that those without health insurance are much more likely to live in strongly Democratic Congressional districts. Many of those contain large minority populations with relatively low incomes.

Texas  is the #1 example.  It would be good to know more.  These are stark differences, not marginal differences in insurance coverage and voting at the Congressional district level.  (map)

Update #1: looking at the map (which is for the population with < 200% poverty line income), there is lots of detail that is non-obvious.  First, there are good and bad states that aren't what you'd expect given their region.  All of New England is good, but NY is bad, especially downstate, but also update (and compare it to the improvement you get when you cross the border to equally depressed upstate Pennsylvania).  New Jersey is bad also.  Alabama and Tennessee are good.  Colorado is bad and Utah is good (well, maybe that's not surprising).  Urban areas are generally worse than rural areas, see for instance Virginia.  Some of this -- how much? -- is simply where recent immigration is, most of it hispanic, see the urban/rural differences and the states: Texas, Florida, New Mexico, California (including the distribution with these states). The midwest does well in part since recent hispanic immigration didn't go there (is this even true?). And again, what is the deal with Alabama? 

Update #2: Jody comments:

There are a ton of Latino immigrants in the upper Midwest now. In a lot of those highly-insured Minnesota counties, the biggest employers are the schools and the county governments. That's a pretty reliable route into insurance. (Also Minnesota has had a robust insurance program for low-income children since the 1980s.)

Yes, there has been a large influx of hispanic immigrants into the Midwest by past standards for that region, but Minnesota isn't a state that has received a lot of this immigration.  States with high hispanic populations in the Midwest are Kansas with 8.9% and Nebraska with 7.5%.  Minnesota is 4.0%, compared to a national figure of 15%.  Connecticut is 11.7%, NY 16.1%.   And then there are places like North Dakota with 1.4% or Maine with 1%. 

My point is that Minnesota still has a small hispanic population.

Going further, I think the high fraction of uninsured recent hispanic immigrants does shape the politics of this issue, see the starting quote of this post.  In some places this is a 'us vs. them' issue, see the infamous 'you lie' outburst thy Rep. Wilson.  

A quick google got me these two figures. It would be interesting to know what these figures look like once one accounts for age, income, occupation, industry, job tenure etc.  I'm surprised that hispanics do so much worse than blacks.  Is it higher public employment by blacks, while hispanics are in traditional non-public industries (agriculture, construction, service work).  And this incidence of non-insured statues does, I think, help explain some of the politics here, at least the Republican opposition. 

There's also a very steep gradient in insurance status by citizenship: 12.9% of native born citizens are not insured, 18.% of naturalized citizens, and 44.7% of non-citizens. 

Great Depression added 6.2 years to life expectancy

José A. Tapia Granadosa  and Ana V. Diez Roux,  Life and death during the Great Depression (PNAS). 

Not that this is the last word on the topic, but I'm glad somebody noticed and wrote a paper, since I've been puzzling about this myself.  (via Razib)

Part of the abstract reads:

Population health did not decline and indeed generally improved during the 4 years of the Great Depression, 1930–1933, with mortality decreasing for almost all ages, and life expectancy increasing by several years in males, females, whites, and nonwhites. For most age groups, mortality tended to peak during years of strong economic expansion (such as 1923, 1926, 1929, and 1936–1937). In contrast, the recessions of 1921, 1930–1933, and 1938 coincided with declines in mortality and gains in life expectancy. The only exception was suicide mortality which increased during the Great Depression, but accounted for less than 2% of deaths.

Update #1: Jody comments:

Workplace mortality? Those were eras of industrial malfeasance.

6.2 years would be a lot of workplace mortality. My impression from the literature is that alcohol and tobacco consumption play a big role here. Still, the effect seems too large and too uniform across demographic groups. 

Human Development Index

This year's human development report is out.  I've never paid much attention to the human development index since it is a aggregate of more interesting subindexes and though I like understanding the factor structure of the underlying subindices, nobody seem to discuss this in the HDR media rumble. 

In any case, how did the US make #13 in the word?  Is that a new development?  Measurement problems with high school dropouts?  No crime statistics?  What happened?  Germany is #22, which I assume is at least in part due to East Germany.  What would West Germany by itself be?  Having the UK ahead of Germany is scary -- though the UK seems to have improved at lot in the last two decades.  Germany should be competing with France (#8) and the Netherlands (#6), not the UK (#21) and Greece (#25). Not sure what fraction of this is due to reunification. 

Low formal schooling enrollment is probably Germany's largest issue in these statistics, #37 in the world. 

Incarceration Risk

Looking through the high school dropout literature to understand what is going on in California, I came across this impressive statistic for the probability of ever going to prison by age 30-34  by race and education.  From Bruce Western, I should find a better cite though.  This is 1999 data, most likely worse by now.  The key figure is that 60% of black male high school dropouts have been in prison by age 30-34. 40% of black males don' t graduate from high school. The incidence this has for the next generation is also impressive. [Update: a podcast of a talk by Bruce Western on this topic, with slides.]

 

California in Crisis

The Guardian writes:

Los Angeles now has a poverty rate of 20%. Other cities across the state, such as Fresno and Modesto, have jobless rates that rival Detroit's. In order to pass its state budget, California's government has had to agree to a deal that cuts billions of dollars from education and sacks 60,000 state employees. ... California's education system has become so poor so quickly that it is now effectively failing its future workforce. The percentage of 19-year-olds at college in the state dropped from 43% to 30% between 1996 and 2004, one of the highest falls ever recorded for any developed world economy. California's schools are ranked 47th out of 50 in the nation. 

Now, incredibly, California, which has been a natural target for immigration throughout its history, is losing people. Between 2004 and 2008, half a million residents upped sticks and headed elsewhere. By 2010, California could lose a congressman because its population will have fallen so much.

I recently noticed that California's share of the US population stopped growing in the mid-1990s and now seems -- how much to the published data based on observations and how much on projections? -- to be falling.  I should dig out the data. 

Also, what data is the Guardian using for college enrollment? Must be from the California Postsecondary Education Commission, but I cannot find the historical time series data for 19 year old freshman enrollment there.  The freshman enrollment rate for HS graduates was 45.3% in 2008, where "college-going rates were calculated by dividing the number of entering freshmen aged 19 and younger from public high schools .. by the total number of graduates from public high schools." Not sure if this is the same concept as quoted above since it only applies to public HS graduates.  

Ah, for "College-Going Rates of Public High School Students Students of All Ethnicities Statewide to Any California Public College or University" the commission posts data from 1989 to 2008.  The high is 50% in 1989 and the low is 43% in 2003.  2008 is 46% and 2007 was 48%.  For the two years cited by the Gardian above we have 49% in 1996 and 45% in 2004.  More volatility here than I would have expected, but less an obvious secular downward trend than what looks like business/budget cycle movement.    Obviously HS graduation rates could be falling and driving the overall college enrollment rate down.  I would need to find that data. Also not sure how to reconcile the 45.3% freshman enrollment rate for 2008 in the above paragraph with the 48% enrollment rate for 2008 in this paragraph, since the higher of the two rates seems to have the same or a narrower definition. 

So to the extent there are large movements in college enrollment rates, they need to be coming form high school graduation rates.  Which is, surprisingly, a whole can of worms in terms of the data, with most official sources being unreliable with upward biases, though there are efforts to clean this up. See for instance Heckman and LaFontaine, The American High School Graduation Rate: Trends and Levels.  

This paper uses multiple data sources and a unified methodology to estimate the trends and levels of the U.S. high school graduation rate. Correcting for important biases that plague previous calculations, we establish that (a) the true high school graduation rate is substantially lower than the official rate issued by the National Center for Educational Statistics; (b) it has been declining over the past 40 years; (c) majority/minority graduation rate differentials are substantial and have not converged over the past 35 years; (d) the decline in high school graduation rates occurs among native populations and is not solely a consequence of increasing proportions of immigrants and minorities in American society; (e) the decline in high school graduation explains part of the recent slowdown in college attendance; and (f) the pattern of the decline of high school graduation rates by gender helps to explain the recent increase in male-female college attendance gaps.

I don't know what good data would show for California. 

Prision Population

via the WSJ:

The  ranking, put out by the King’s College London International Centre for Prison Studies, found that in the U.S. 756 out of every 100,000 is incarcerated.

According to the study:

The United States has the highest prison population rate in the world, 756 per 100,000 of the national population, followed by Russia (629), Rwanda (604), St Kitts & Nevis (588), Cuba (c.531), U.S. Virgin Is. (512), British Virgin Is. (488), Palau (478), Belarus (468), Belize (455), Bahamas (422), Georgia (415), American Samoa (410), Grenada (408) and Anguilla (401).

We placed a call to Douglas Berman, the Ohio State law professor who runs the very fine Sentencing Law and Policy blog, to ask for help on making sense of the numbers.

Berman sees the statistics as, in a word, ironic. “The chest-pounding about freedom and liberty that so many of our leaders do, both on the left and the right seems inconsistent with this statistical anomaly,” he says. “We live in a country conceived on the notion of liberty, but we lead the world in locking people up.”

Germany is at 89 per 100,000 or 11.8% of the US rate.  Canada at 116 per 100,000 or 15.4% of the US rate. 

Occupations with the most auto insurance claims

via MaximumPC, a list of occupations with the most auto insurance claims:

1. Computer Engineer
2. Sales Manager
3. Chef
4. Student
5. Doctor
6. Estate Agents
7. Surveyor
8. Customer Advisor
9. Hairdresser
10. Social Worker

Fewest claims:

1. Farmer
2. Aircraft Fitter
3. Stores Personnel
4. Ambulance Driver
5. Laboratory Technician
6. Pilot
7. Caretaker
8. Agricultural Engineer
9. Green Keeper
10. Mechanical Engineer

A lot of this may simply reflect car values.  At least for doctors. 

House Prices and Obesity

From a study of Seattle area obesity -- not that I necessarily would trust MDs with statistics.  Still, the correlations seem astoundingly high for a community level predictor.  How does this happen?  Doing multivariate analysis with household income and housing price as well as demographics would be interesting. 

The goal of this study was to determine whether obesity rates in King County, Seattle, Washington state, at the ZIP code scale were associated with area-based measures of socioeconomic status and wealth. Data from the Behavioral Risk Factor Surveillance System were analyzed. At the ZIP code scale, crude obesity rates varied six-fold. In a model adjusting for covariates and spatial dependence, property values were the strongest predictor of the area-based smoothed obesity prevalence.

Predictor	r2	p-value
Median House Value	0.46	<0.001
Median Household Income	0.05	0.058
% of Adults Living Below Poverty Line1	0.02	0.26
Percent African-American	0.02	0.18
Percent Hispanic	0.19	<0.001

 

Figure 1

Crude obesity prevalence and 95% confidence intervals for ZIP code areas in King County

(link)

The Unequal Geographic Burden of Federal Taxation

Since I'm moving to an even higher wage and higher cost location, I though I'd mention this. The point -- despite what the title of the article may imply -- isn't that redistribution from higher income to lower income persons is bad, but that who is higher income in real terms is mismeasured by the US tax system and that this mismeasurement has efficiency costs. Do other countries do better on this dimension?

David Albouy, The Unequal Geographic Burden of Federal Taxation

In the United States, workers in cities offering above‐average wages—cities with high productivity, low quality of life, or inefficient housing sectors—pay 27 percent more in federal taxes than otherwise identical workers in cities offering below‐average wages. According to simulation results, taxes lower long‐run employment levels in high‐wage areas by 13 percent and land and housing prices by 21 and 5 percent, causing locational inefficiencies costing 0.23 percent of income, or $28 billion in 2008. Employment is shifted from north to south and from urban to rural areas. Tax deductions index taxes partially to local cost of living, improving locational efficiency.

Ethnic Differences between CT and VA Census Tracts

Since I was in the 2000 census data, I looked at the ancestry tables.  They seem to only exist for whites.  So, restricted to that subgroup, there are significant differences between the two census tracts, something I'd not thought about but which makes sense.  Basically, we loose Italian (-12%), Polish (-3.3%) and Russian (-6.5%) ancestry and pick up English (+11.5%), German (+10%) and Irish (+5.5%).  Which may reflect moving away from the urban Northeast and into 'the South', though I'd hesitate to call northern VA part of the South.  It may also mean fewer Jews, but there's no direct data on that.  Is there a Jewish area in Northern VA?   Some shifts in ethnic composition are pretty extreme, for instance Russian ancestry going from 7.65% to 1.23%.  Germans go from 6.1% to 15.7%, English from 6.4% to 18%. Irish from 11.7% to 17.3%.  Combined English, German and Irish goes from a tad under 25% to a bit over 50%. 

A town with no synagogues. 

What is weird: French Canadian ancestry doesn't budge, moving for 1.66% to 1.65%, and that doesn't look like sample noise.  What's going on here?  I think of French Canadians as a Northeastern and maybe midwestern ethnic group that has done quite well and should be in the suburbs. And indeed, they are, since nationally French Canadians are only 0.82% of the population. [Update #1: however, there is also a 'French' ancestry that seems to be a exact equivalent of 'French Canadian' and which is almost 3% of the US population...more later.  Hey, who knew there were around 1/3 as many French Canadians as blacks in the US.  More French Canadians as all Asians put together.  Puts things in perspective.]

Also, there is about 1.9% British -- in addition to English -- ancestry in VA, none in CT.  Similar, 2.65% Ukrainian -- in addition to Russian -- ancestry in CT, 0.68% Ukrainian ancestry in VA, both reinforcing the patterns noted above.  Same for Scottish ancestry.

One more puzzle: the CT tract shows  

There may have been some change since 2000. 

Update #1: finally, the racial composition of the CT and VA census tracts, at least for 2000, did surprise me.  I'd not expected we'd loose racial diversity moving to VA, since so far northern VA is reminding me of my stint at the World Bank, lots of professional and lots of Asians, especially South Asians.  But this isn't what this census tract shows. And who knows where we'll end up. 

                           US        CT tract     VA tract

non-hispanic white	69.13%	88.47%	92.01%
Hispanic	12.55%	1.72%	2.65%
Black	12.06%	1.76%	1.38%
Asian	3.60%	6.06%	3.15%

Educational Segregation

So we're selling a house in CT and trying to buy a house in VA.  So what do I do on a Sunday morning?  Look up educational demographics by census tract, after looking at houses yesterday and facing more house searching during the week.  One issue that comes up is general 'neighborhood' characteristics.  I liked one area I saw for the first time yesterday and it has good reputation.  Nothing that looks over the top for the most part, unlike some other blighted areas.  Real trees and (somewhat) winding roads, but also very upper middle class suburban, built in the 1970s.  In other words, like the sort of place we live now.  

So here are the relevant census tract educational attainment stats for people 25+ years old, for the nation at large, our current census tract and the VA census tract I'm interested in.  The two tracts have a bit under 3,800 and a bit over 2,000 people 25+ years old. 

                       US           CT tract    VA tract

Men

some college or less	73.90%	33.09%	12.70%
college	16.11%	21.21%	28.38%
college+	9.99%	45.70%	58.91%
Ph.D	1.38%	10.07%	9.02%

Women

some college or less	77.15%	45.80%	37.83%
college	15.02%	25.41%	29.94%
college+	7.83%	28.79%	32.22%
Ph.D	0.58%	5.37%	4.66%

These are not small demographic differences to the US at large.  What people mean when they talk about coastal elites?  

Majority-Minority Areas in China

Fallows writes:

The NYT online has a very nice graphic just now showing the parts of China with significant "minority" population. Minority, in this sense, means one of the 55 recognized groups other than Han Chinese that together make up about 8 percent of the country's population. The screen shot below is not the default version of the graphic, which shows all counties in China with at least 10 percent minority population. Instead it's the version that shows counties where at least half the people are something other than Han.



In a sense the map is misleading, in the same way "Red State / Blue State" electoral maps are misleading about real division of opinion within the United States. The big western areas marked as Tibetan or Uighur are very lightly populated (think Nevada, Montana), compared with the dense, mainly-Han areas of the east. For instance, the ethnic Tibetan areas are shown as covering not just Tibet proper but also parts of the neighboring provinces of Sichuan, Qingai, and Gansu (in all of which places I have been to Tibetan villages). But the total ethnic-Tibetan proportion of China's population is something under one percent. Still, the graph gives an idea of the control issues China has with some of its minority groups.

In terms of population these don't seem to be large groups, though you'd need to know how the Chinese state classifies people's ethnicity.  On the other hand, the map above is most likely based on the same Chinese state sources. Wiki writes:

The People's Republic of China (PRC) officially recognises 56 distinct ethnic groups, the largest of which are Han Chinese, which constitutes about 91.9% of the total population. The 55 other ethnic groups are officially recognised as ethnic minority groups. The large ethnic minority groups in terms of population include the Zhuang at 16 million, the Manchu at 10 million, the Hui at 9 million, the Miao at 8 million, the Uyghur at 7 million, the Yi at 7 million, the Tujia at 5.75 million, the Mongols at 5 million, the Tibetans at 5 million, the Buyei at 3 million, and the Koreans at 2 million.

Gender Gap in Undemployment Rates

via Clusterstock.  This looks like it might have an interesting story behind it, since it doesn't seem all that stationary, or rather stationary with occasional breaks.  Two big breaks there, 1979-83 and now, with smaller surges during other recessions that then reverted. 

Chart is male unemployment rate minus female unemployment rate. 

The Collapse of (parts of) the Oregon Economy

The NYT puts the story on the front page with the usual up close and personal anecdotes.  Still, the important observation that this is high unemployment in a high growth state with lots of immigration in the last decade. Datelined in Bend, about 80 miles east of Eugene with a population of 80,000 now, up from 50,000 in 2000, the NYT writes:

Now the Californians who contributed to Oregon’s growth are in some cases adding to its economic struggle. As of May, Oregon had the second-highest unemployment rate in the nation, at 12.4 percent, behind Michigan. California, which has not released its May figures, ranked fifth in April.

While some other states with high unemployment, including Michigan, have seen their labor forces shrink, Oregon’s labor force has grown. Economists say some of the growth appears to be driven by people who moved here with money they made in California, whether from real estate or stock market investments, and expected to get by but now must look for work.

“It’s just so depressing to hear them because they thought they had life handled and they don’t,” said Bobbie Faust, an employment counselor who works for the state in Bend.

The Telfords are among those facing trouble. They had presumed they would be able to sell their house in Fresno for more than $300,000 to help pay the mortgage on the new house they bought near the Deschutes River in Bend for $475,000. But the Fresno house has yet to sell, and Mrs. Telford, an accountant, has lost a series of jobs at small firms here that she said had downsized. The couple’s only income now comes from her unemployment checks and her husband’s salary as a high school teacher.  ....

Not all of the newcomers are from California, of course. Lost equity, lost jobs and the possibility of foreclosure also threaten people who moved here from just across the Cascade Range, on the wetter western side of Oregon, as well as some from Seattle or the East. Measuring California’s economic impact on Oregon and its struggles is difficult, and economists say that Oregon, which has less than a tenth of the population of California, has not always been directly affected by its neighbor’s fortunes.

Still, just as other places in the West have blamed California transplants for treading heavily into town, the words “California equity” roll off many tongues here in Deschutes County with particular resentment these days.

“California immigrants can never win in Oregon,” said Philip J. Romero, an economist who has advised governors in both states. “In a boom, ‘They are crowding the roads and bidding up house prices.’ In a bust, it’s: ‘They alone caused the price of my house to drop by hundreds of thousands of dollars. They came up here without a job, and now we can’t absorb them and they’re competing for my job.’ ”

Carolyn Eagan, a regional economist for the Oregon Employment Department, pointed to federal data showing that the overall percentage of personal income from dividends, investments and rental income in Deschutes County was almost 26 percent in 2007, the latest year for which data were available. Compared with an overall state rate of slightly less than 21 percent, the figures suggest that people here, more than elsewhere, have relied on income from sources other than a steady job.

There are homes in Fresno for $300,000 owned by middle class dual income households? 

Finally, via wiki:

Tourism is one of Bend's largest sectors. The Mount Bachelor ski resort brings in tourists from all over Oregon, Washington, and California. The nearby Cascade Lakes are also a large draw for tourists. Recreational activities include downhill and cross country skiing, hiking, biking, rafting, golfing, camping, fishing, picnicking, rock climbing, and general sightseeing.

Bend is also home to the Deschutes Brewery, the largest of several microbreweries in the city. Since 2004, Bend has also hosted the Bend Film Festival.

In 2005, Bend's economic profile comprised five industry categories: tourism (7,772 jobs); healthcare and social services (6,062 jobs); professional, scientific and technical services (1,893 jobs); wood products manufacturing (1,798 jobs); and recreation and transportation equipment (1,065 jobs).

...in 2005 construction and real estate accounted for 17.3% of all jobs in the Bend metropolitan statistical area (MSA)...

Much of Bend's rapid growth in recent years is also due to its attraction as a retirement destination. The rapid population growth has fostered organizations such as Central Oregon Landwatch and Oregon Solutions

....

In June 2006, Money magazine named the Bend MSA the fifth most overpriced real estate market in the United States. By September 2006, the Bend metro area ranked second in the list of most overpriced housing markets, and in June 2007 it was named the most overpriced housing market in America.

Update #1: David comments:

We just visited Bend this past Winter. It's beautiful. But we did notice lots of fancy boutiques with big clearance sales in downtown. No waiting for tables at restaurants and the ski condos offer big discounts.
The skiing there is world-class and it's cheaper and less crowded than Tahoe or Vermont (for example).
I like the California migration, since many of them are old enough to have Medicare. On the other hand, I've seen lots of patients recently who are about to loose their insurance and panicking. Several patients have refused to come back for routine office visits because of cost. They'd rather check their own vital signs and call me for instructions, saving the charges of having to come in to the office.
It does feel a little strange here. We personally know about 5 parents of our kids' friends who have been laid off. I had a patient yesterday who got laid off from her job as a clerk for the state government after 20 years and she's now about to loose her house.

'Has the Rise in American Inequality Been Exaggerated?'

Robert Gordon has a new paper out, Has the Rise in American Inequality Been Exaggerated?  I like Gordon's point that rising housing costs have impacted college educated households more non-college educated households.  I've wondered about this myself, in particular in the context of understanding how it affects college vs. non-college life cycle choices about housing and fertility. Gordon is no right wing economist, just look at his final Political and Policy Implications section. 

The rise in American inequality has been exaggerated both in magnitude and timing. Commentators lament the gap between the growth in real median household income and the much faster growth in productivity. This paper shows that a conceptually consistent measure of this growth gap over 1967 to 2007 is only one‐sixth of the conventional growth gap. Further, the timing of the rise of inequality is often misunderstood. The lamentably slow growth in real median household income in 2000‐ 07 is not due to higher inequality, and indeed inequality decreased slightly after 2000. This cessation of inequality’s secular rise in 2000 is evident from the growth of Census mean vs. median income, and in the income share of the top one percent of the income distribution. The income share of the 91st to 95th percentile has not increased since 1983, and the income ratio of the 90th to 10th percentile has barely increased since 1986. Further, despite a transient decline in labor’s income share in 2000‐06, by the end of 2008 labor’s share had returned to the same value as in 1983, 1991, and 2001.

Recent contributions in the inequality literature have raised questions about previous research on skill‐biased technical change and the managerial power of CEOs. Directly supporting our theme of prior exaggeration of the rise of inequality is new research showing that two‐thirds of the post‐1980 increase in the college wage premium disappears when allowance is made for the faster rise in the cost of living in cities where the college educated congregate and for the lower quality of housing in those cities. A continuing tendency for life expectancy to increase faster among the rich than among the poor reflects the joint impact of education on both economic and health outcomes, some of which are driven by the behavioral choices of the less educated.

More Retrospection on the 2008 Election

Nate Silver:

In 1980, 32 percent of the electorate consisted of [non-hispanic] white Democrats (or at least white Carter voters) -- likewise, in 2008, 32 percent of the electorate consisted of [non-hispanic] white Obama voters. But whereas, in 1980, just 9 percent of the electorate were nonwhite Carter voters, 21 percent of the electorate were nonwhite Obama voters last year. Thus, Carter went down to a landslide defeat, whereas Obama defeated John McCain by a healthy margin.

Andrew Gelman:

[L]et's try the counterfactual in which the Republicans give up on the Hispanic vote, which I'll operationalize by transferring half of McCain's Hispanic votes in every state to Obama. (For example, we estimate that McCain got 22.8% of the two-party vote among Hispanics in New Jersey. Under this counterfactual, we'll give him just 11.4%.) Here's what happens:

Again, not much difference. Ummm . . . Missouri moves to 50.1% for Obama. And here's the scatterplot:

The bottom line: Hispanics were not a key component in Obama's win. However, this is not to say that the Republicans should not try to contest the Hispanic vote. As the last scatterplot above shows, further losses of Hispanics would make the Democrats competitive in Georgia, Texas, and Arizona. In some sense this is no big deal, at least at the presidential level: If the Democrats remain at 53% or 54% of the vote, they'll win nationally in any case. If we imagine a national swing of 3% or so toward the Republicans, so they're competitive nationally, then their big risk if they lose Hispanic votes is to no longer be viable in Florida (where we estimate McCain to have won 43% of the two-party vote among Hispanics in 2008). That's the state where Republicans really can't afford to abandon the Hispanic vote.

P.S. Some commenters point out that the Hispanic vote is expected to [grow]. Following up on the above, I did some crude calculations, assumning that the Hispanic vote share increases by 20% in each state:

Again, the bottom line is that the biggest difference is in Florida, with its high Hispanic vote that is currently nearly evenly split between the two parties. Texas and Arizona show big potential shifts too, but, again, if these states are swinging without other big changes happening elsewhere, the national Republican party is in big trouble anyway.

Do Republican Administrations cause out of wedlock child births?

The not terribly statistically significant aggregate timeseries data sure look like it, with big surges in the unmarried mother birth rate both under Reagan and Bush II, and no increase, despite a secular upward trend, under Clinton and a only a small increase under Carter. Knowing more would be nice. And I have been annoyed by the whole unwed teenage motherhood of Bristol Palin -- Republican can still claim to be 'socially conservative' while presiding over large expansions in the number of unwed mothers compared to the record of Democrats and having glamorous unwed teen mothers in the family of their presidential candidates.  And yes, I do think lots unwed mothers are, given the way things play out, a social issue that ought to be addressed (along the appropriate margins). 

Update #1: the above is written looking at the unwed birth rate.  If you look at the fraction of unwed births the curve is much smoother (probably an indication that this is the right graph to look at), there are only three big facts:

1. a secular upward trend starting somewhere around 1960 (unwed births head up earlier, but that is just the overall birth rate heading up -- still, it seems odd that unwed and wed births would be driven by the same forces).
2. a near stop in this upward trend somewhere around 1992/93 (too early to be due to Clinton?).  It would be really really nice to know what is causing this big break in the timeseries. 
3. the resumption of the upward trend in the first Bush II term. 

Graph from NCHS.  Data look a bit weird too, so I ought to check the underlying data....

Disassociative Mating

This is cool (via Paul Kedrosky):

Scott Rick, Deborah A. Small, Eli Finkel, Fatal (Fiscal) Attraction: Spendthrifts and Tightwads in Marriage

Although much research finds that "birds of a feather flock together," surveys of married adults suggest that opposites attract when it comes to emotional reactions toward spending. That is, "tightwads," who generally spend less than they would ideally like to spend, and "spendthrifts," who generally spend more than they would ideally like to spend, tend to marry each other, consistent with the notion that people are attracted to mates who possess characteristics dissimilar to those they deplore in themselves (Klohnen and Mendelsohn 1998). In spite of this complementary attraction, spendthrift/tightwad differences within a marriage predict conflict over finances, which in turn predict diminished marital well-being. These findings underscore the importance of studying the relationships between money, consumption, and happiness at an interpersonal level.

I'm not sure they have the matching process right here.  An alternative view would be there is an ideal level of frugalness for everybody and deviations from that level in one partner need to be paid for by deviations in the level of frugalness in the other partner, irrespective of sign.  Good people match up with good people, less good people with less good people. Does this go beyond this?  Do excessively frugal people matched up with excessively frugal people do better? 

I couldn't actually say where I and my wife line up here.  We think of ourselves as similar in frugalness, but who knows. 

Patna and the Maurya Empire

Looking at large pre-industrial cities in the world (see below), Patna shows up in 200 BCE as the 2nd largest city in the world with a population of around 350,000, right after Changan (i.e. Xian in China, capital of China for most of the time until the 10th century CE.) with a population of 400,000.  Both cities are bigger than any previous cities in history and in the same size category as Rome at its largest, with only Baghdad significantly surpassing this city size in the 9th century CE. 

I'd not realized where Patna is (Bihar, on the southern bank of the Ganges) nor that it was the capital of the Maurya Empire, which lasted from 321 to 185 BCE  and which was larger than I'd realized, both territorially -- encompassing most of north India including Pakistan today and at times large parts of south India and Afghanistan --  and populationwise, with figures cited up to 50 million (which may be in part modern overstatements, but even 30 million would be impressive).  The Roman Empire at its height seems to be estimated up to about 60 million, but I'd worry about overestimates by enthusiasts here as well (and what were the incentives for census takers?).  Finally, there is always the danger that one things of the Empire as having more local impact or control than is warranted (but the impact of imperial control can at times be astonishingly strong at times). 

Still, for its size, duration and urbanization, the Maurya Empire seems to get little recognition in popular culture.  Why is this?  Its most famous ruler appears to have been Ashoka the Great, with a reign from 273 to 232 BCE.  Also the first large Buddhist polity. The last Maurya ruler was overthrown by the Hindu Sunga in 185 BCE in the context of military intrusions from Indo-Greek kingdoms in the Afghan to Punjab area. 

Survey Errors

This is a pretty impressive survey error, especially given the relatively non-sensitive nature of the question.  Monkey Cage writes:

In a study reported in the current issue of Public Opinion Quarterly, Markus Prior says ... self-reports of exposure to the evening news shows are grossly off.

That conclusion is based on Prior’s comparison between self-report data from the 2000 National Annenberg Election Survey (NAES), on the one hand, and ratings of news audiences based on Nielsen’s “people meters,” on the other. (In each of 5,000 Nielsen households, each TV set was attached to a meter that household members used to indicate the beginning and end of their viewing.)

Prior’s basic finding:

The survey estimates vastly overstate the size of the network news audience. According to Nielsen, between 30 and 35 million people watched the nightly news on an average weekday. Based on NAES self-reports, that number is between 85 and 110 million for most of the year. In other words, the ANES -based estimate is somewhere around two-and-a-half to three-and-a-half times higher than the Nielsen-based estimate.

Oof.

But it gets worse. The overreport factor varies from one part of the public to another. For example. For viewers aged 55 or more, it’s about two-to-one; but for those in the 18-34 age range, it’s somewhere around six-to-one or even eight-to-one.

Why did interstate geographical mobility fall in 2005-2006?

The census has its latest geographical mobility figures out (via CR). 

The weird thing here is that interstate mobility falls already in April 2005 to March 2006 period (the data is from the March CPS survey). In particular, the data is 

April-March          interstate movers
2007-2008             1.6%
2006-2007             1.7%
2005-2006             2.0%
2004-2005             2.6%
2003-2004             2.6%
2002-2003             2.7%

This is a large fall from 2004-2005 to 2005-2006, with the 2005-2006 level of 2% lower than the previous post-WWII record low of 2.2%.  Note that this is before any big problems developed in the housing market.  Why is this???  Move rates look countercyclical, but with a good amount of lag.  Understanding the mechanism here would be interesting and what that mechanism means for the interpretation of the current move rate. 

Fernando Ferreira, Joseph Gyourko, and Joseph Tracy Housing Busts and Household Mobility argue that

Using two decades of American Housing Survey data from 1985-2005, we estimate the impact on household mobility of owners having negative equity in their homes and of rising mortgage interest rates. We find that both lead to lower, not higher, mobility rates over time. The impacts are economically large, with mobility being almost 50 percent lower for owners with negative equity in their homes. This does not imply that current worries about defaults and owners having to move from their homes are entirely misplaced. It does indicate that, in the past, the lock-in effects of these two factors were dominant over time. Our results cannot simply be extrapolated to the future, but policy makers should begin to consider the consequences of lock-in and reduced household mobility because they are quite different from those associated with default and higher mobility.

But this doesn't explain why 2005-2006 interstate mobility is so low, since negative housing equity is still relatively low and interest rates aren't high (though higher than in the previous years -- is this it?  For such a large drop in mobility?  It would be interesting to see the demographic breakdown of who moves and how it changes over time.

UN Population Projections

It seems the most recent UN Population Projections came out last week.  Interestingly, and I can't quickly determine if this is new, but probably not, the low population growth scenario (there is low, middle and high growth scenarios) has world population reaching a high of 8.024 billion in 2040 -- all my kids then will still be younger than me now, so it is not that far in the future -- and falling slowly thereafter. 

On the other hand, as Joel Cohen points out in the NYT, there is a surprisingly large amount of uncertainty in these estimates:

The projected population of the world in 2050 is 10.46 billion under the U.N.’s high variant and 7.96 billion under the low variant–a difference of 2.5 billion people. This difference equals the total population of the Earth a century earlier, in 1950. The difference between the assumptions of the high and low variants is a difference of one child per woman from now to 2050.

As Longman writes in USA Today, the spread of low fertility is not a OECD country affair, but global:

Yet starting in the world's richest, best-fed nations during the 1970s,and now spreading throughout the developing world, we find birthrates falling below the levels needed to avoid long-term, and in many instances, short-term, population loss. The phenomenon has spread beyond Europe and Asia to Latin America.

Brazil, a land once known for its celebration of dental-floss bikinis and youthful carnival exuberance, is an aging nation that no longer produces enough children to replace its population.The same is true of Chile and Costa Rica. Joining them over the next 10 to 20 years, the U.N. projects, will be many other countries Americans still tend to associate with youth bulges including Mexico, Argentina, Indonesia, India, Vietnam, Algeria, Kuwait, Libya and Morocco. Think we need to build a wall on the southern border? Birthrates have declined so quickly in Mexico that its population of children younger than 15 has been in free-fall since 2000 and is expected to drop by one-third over the next 40 years.

The remaining population growth is for the most part going to be in Africa and among old people. Total world births in the medium growth scenario start falling in the late 2010s, in the low growth variant in the immediate future. The total number of young people starts falling in line with births with the expected age determined delay. 

Mexico's total births have been falling since the early 1990s, unlike the US, which is at an all time high for total briths.  Brazil's total briths peaked in the early 1980s and are noiw 15% lower.  Morocco's total births peaked in the early 1980s as well. [Obviously total births aren't going to be single peaked under many paths for the fertility rate, but it looks like they will be for countries undergoing fast demographic transitions to below replacement rate fertility.]

Also interesting, world population aging, now rapid, slows down considerably after 2050 according to projections. 

So when did the UN start predicting decling world population and how has the timing for the peak changed? 

Update #1: and most importantly, only a bit over 1.1 billion people more by the peak under the low growth scenario, from a bit under 6.9 billion now to a bit over 8 billion in 2040.