The World's Most Expensive Bridge Toll

It is quite remarkable that a bridge can support such high traffic levels with such a high toll, which implies that even with the high toll, it is superior to the alternatives in the eyes of people who use it.

A bridge connects Copenhagen [in Denmark] and Malmo [in Sweden], and now the price is higher:

…the basic price for a one-way car journey across the bridge has been jacked up to 510 Danish kroner, or £58. For the largest vans, it is the equivalent of £218.

Research by Sydsvenskan, a regional newspaper in southern Sweden, suggests this is by far the most expensive bridge toll on the planet, costing about twice as much as its nearest rivals in Japan and Canada…

Despite the vehicle toll, the total number of people crossing the Oresund by car, train or ferry hit a record 38 million last year, equivalent to about 105,000 trips a day. A one-way railway journey between central Copenhagen and Malmo typically costs only £13.

Via Marginal Revolution. 

Predictors Of Long Term Unemployment In Sweden

The availability of huge government collected data sets facilitated by a well functioning government and public trust that allows privacy concerns to be overcome due to a history of non-corruption makes Scandinavia one of the best places in the world to do large sample size social science research. This makes this research valuable despite the great differences in culture and economic organization between places like Sweden and societies like the United States.

In a nutshell, a recent paper on long term unemployment in Sweden concludes that some individuals who are long term unemployed are at much greater risk of ending up that way, especially in economic downturns, than others. This is best predicted by their past history of unemployment.

This paper studies the predictability of long-term unemployment (LTU) and analyzes its main determinants using rich administrative data in Sweden. Compared to using standard socio-demographic variables, the predictive power more than doubles when leveraging the rich data environment. The largest gains come from adding job seekers' employment history prior to becoming unemployed. Applying our prediction algorithm over the unemployment spell, we show that dynamic selection into LTU explains at least half of the observed decline in job finding. While the within-individual declines are small on average, we find substantial heterogeneity in the individual-level declines and thus reject the commonly used proportional hazard assumption. Applying our prediction algorithm over the business cycle, we find that the cyclicality in average LTU risk is not driven by composition but rather by within-individual cyclicality and that individual rankings are relatively persistent across years. Finally, we evaluate the implications of our findings for the value of targeting unemployment policies and how these change over the unemployment spell and the business cycle.

Andreas I. Mueller and Johannes Spinnewijn, "The Nature of Long-Term Unemployment: Predictability, Heterogeneity and Selection" NBER WORKING PAPER 30979 (February 2023) DOI 10.3386/w30979.

IQ v. Income In Sweden

Are the best-paying jobs with the highest prestige done by individuals of great intelligence? 

Past studies find job success to increase with cognitive ability, but do not examine how, conversely, ability varies with job success. Stratification theories suggest that social background and cumulative advantage dominate cognitive ability as determinants of high occupational success. This leads us to hypothesize that among the relatively successful, average ability is concave in income and prestige. 

We draw on Swedish register data containing measures of cognitive ability and labour-market success for 59,000 men who took a compulsory military conscription test. 

Strikingly, we find that the relationship between ability and wage is strong overall, yet above €60,000 per year ability plateaus at a modest level of +1 standard deviation. The top 1 per cent even score slightly worse on cognitive ability than those in the income strata right below them. We observe a similar but less pronounced plateauing of ability at high occupational prestige.

Marc Keuschnigg, Arnout van de Rijt, and Thijs Bol, "The plateauing of cognitive ability among top earners" European Sociological Review, jcac076 (January 28, 2023) (open access). Hat tip to Marginal Revolution.

The introduction also notes that:

[I]ncome distributions have strong right skew. In all Western countries, top income shares have been steadily rising since the 1980s, with the 1 per cent highest earners receiving 9 per cent of national income in Sweden and even 20 per cent in the United States—excluding capital gains[.]

Note the subtlety that while very high levels of IQ translate on average to continued improvement in economic and non-economic outcomes, that lots of people who achieve high levels of economic and non-economic outcomes are people of lesser IQ who are at the top of the bell curve for outcomes for a given IQ. 

Within the bottom 50% of economic outcomes the correlation of IQ and economic results, rather than being monotonic, hits minimum IQ at the 20th percentile of income and starts rising below that point.

The range between the 50th and 90th percentiles of economic outcomes seems to be the most meritocratic.

Then, there is a leveling off in the top 10%.

You might reasonably have guessed in the absence of data, that the top 10% of economic outcomes would be extremely meritocratic because lots of jobs paying in the top 10% are highly skilled in a way that imposes a de facto minimum IQ, while "faking it" might be easier in middle class jobs that are less skilled. But that doesn't seem to be the case until you get to jobs in the bottom half of the wage distribution.

On the other hand, as someone who regularly deals with lots of people in the top 1% of income as a lawyer, I am well aware that they are not predominantly at or close to the top 1% of IQ, although they are rarely entirely stupid either. The epithet used to be that: "A students work for B students in companies owned by C students", although grade inflation has distorted that a bit over time. 

A plateau at €60,000 per year is lower than I would expect, although given the relatively egalitarian economy and small scale of the Swedish economy, it is somewhat harder to put a finger on what is going on as it differs in lots of material ways from the U.S. economy with which I am much more familiar. 

There is also no doubt that non-IQ personality factors like "work ethic", conscientiousness, and extraversion, play a part in economic outcomes as well. But, I'd would be surprised if those factors eclipsed IQ at the 10th percentile of income and would be less surprised if the administrative and professional sector of Sweden was less meritocratic than lower level employment.

For example, I have only a dim idea of how important family wealth is to adult income in Sweden, even though I'm much more familiar with this in the U.S.

It is also possible that somewhere in the top 10% that wage data ceases to be an accurate measure of economic returns to ability, with profits from businesses that you own and royalty income, for example, becoming more relevant than mere wages for this group of workers. In the U.S. this tendency is driven, in part, by tax law in addition to the fundamental economics, in ways that I understand fairly well, but again, I don't know how this plays out in Sweden and don't trust my intuition about a situation that in the U.S. that is heavily driven by U.S. tax laws to generalize to Sweden which has very different tax laws.

The paper's own theoretical prediction suggests that the actual data best correspond to a situation in which there is an extremely "noisy" relationship between IQ and income:

The sample size is 59,000 people, which is decent, but the top 1% is still only about 590 men (the study is based upon military entrance exams so it doesn't address income to IQ relationships for women). Also, top income may partially be suppressed by the fact that the cohort of men involved have not yet reached their peak earnings years. The sample is as follows:

Our analysis includes men who joined the labour force between 1991 and 2003 (median 1993). In total, 670,203 men, aged 18–60, entered the labour market in this period to be fully employed for at least 1 year. Cognitive-ability scores are available only for Swedish men who had the obligation to undergo military conscription. We subset on men who took a compulsory conscription test at age 18–19 during 1971–1977 or 1980–1999, years in which ≥90 per cent of each cohort enlisted (94 per cent on average). Enlistment became less comprehensive after 1999 and dropped substantially until its abolition in 2010. We focus on multi-year career success for those men for which we observe 11 years of labour market participation centred around the age of 40 (see below). This leaves us with a full census of 59,387 Swedish-born men for whom we observe a balanced 11-year panel of annual labour-market success.

Cognitive ability is on a scale of 1-9 with a median of 5, a mean of 5.45, and a standard deviation of 1.98. About 30% of the sample had scores of 7 and above, and about 15% has scores of 8 and above. So the plateau in the top 10% of income earners embraces around the top 25% by IQ.

The distribution of wages has a long tail at the high end.

The distribution of the underlying quantities compared is as follows:

The comments to the Marginal Revolution post from a person identified only as "M" discusses the related issue of higher incomes in urban areas notes some sources for further reading:

https://www.eurekalert.org/news-releases/977543 

- "Benefits of big city life – only for the elite - Urban scaling laws arise from within-city inequalities"

and previously 

https://www.eurekalert.org/news-releases/629397 

- "Big cities feed on their hinterlands to sustain growth" - "Their research, published in Science Advances, shows that big cities feed on their hinterlands to sustain growth, thereby escalating the urban-rural divide in economic prosperity and individual life chances. Individuals who leave small areas for large cities are better educated and have higher cognitive abilities than those who stay. These findings provide a more nuanced account for the reasons behind the increasingly uneven economic geography observed in many countries, with growing levels of inequality between urban and rural areas."

Urban superlinear scaling in large cities mostly caused by, and benefits to, high talented individuals. Not a lossless phenomenon for the rest of the country, so extending largest cities, although not zero-sum, generates fewer positive sum gains than thought (contra "Extending the most productive cities would be the way to most strongly boost national living standards").

Marital Child Bearing Rates And Divorce Rates Are Still Driven By Economics

None of this is new to regular readers of this blog, but since I wrote this recap in a Facebook post and comment, where it was really too long, I am going to repost this analysis here with some minor stylistic changes (e.g. you can't do paragraph breaks in Facebook comments).

The Post and Comment

[A]mong babies born to mothers who never went to college, a clear majority are born out of wedlock (58 percent). . . . Among babies born to mothers who went to college but never earned a degree, the out-of-wedlock rate is 44 percent. It’s only 10 percent for college-educated mothers.

From an opinion piece in the Washington Examiner (a conservative D.C. periodical). 

Notwithstanding the real thing called the "Anti-Flynn Effect" that has shown up in a few countries, I think I know what is driving this trend and it is economics, not intelligence. 

Real wages for non-college men have economically stagnated since the 70s and they face intermittent unemployment, and they haven't moved into formerly "women's" jobs. 

Economic prospects for non-college women have soared since the 1970s as new job fields have opened up and braun has become less economically important, and taking time off to have kids involves fewer long term economic sacrifices for less skilled work (there isn't a seniority/experience premium). 

Real wages for college educated people have surged since the 1970s absorbing almost all economic growth. 

College educated women who take time off to have kids pay a huge long term economic price for doing so since there are huge experience/seniority/maximal hours per week premiums in highly skilled work like law, public accounting, engineering, medicine and senior management in businesses, even though college women with children still have better economic prospects than non-college women with children. 

The relative economic potential of non-college men relative to non-college women with children has plummeted in the exact time that non-marital child bearing for them has surged, while the relative economic potential of college men relative to college women with children has stayed high. 

Women are willing to have children with men who are great providers relative to them, but the benefit of marriage for women is to capture the economic potential of their partners which they are dependent upon and that's grown pretty meager for non-college women. 

You see the same decline of marriage in places like Sweden where the long term economic sacrifice college educated women make to have children is much smaller.

And, you see very high marriage rates among women who have children, in places like Japan where the long term economic sacrifice women (even non-college educated women) who have children face is very high. 

African-American marriage collapsed first because African-American men have higher unemployment and lower wages for comparable levels of education than whites, plus the handicap of mass incarceration at far higher rates, while African-American women are at a much smaller economic disadvantage relative to white women, and because fewer African-Americans are college educated. 

A lot of conservative anti-gay rights sentiment is driven by the fact that their marriage institution is falling apart but they don't understand that the cause is mostly economic. A lot of conservative "anti-woman" sentiment is driven by the fact that greater economic prospects for non-college women is half the reason that their marriages and families are weaker (the other half being the stagnation of working class economic prospects due to automation, offshoring, immigration, and men's unwillingness or inability to equip themselves for an economy that now needs more brain than brawn. 

African-American communities drew different political lessons from the same phenomena because they were willfully excluded from conservative political circles by white racism and because the economic and incarceration causes of the decline of marriage in their communities was more obvious to them.

Further Discussion

It is worth noting how compelling the desire to have kids and spend some time as a primary caretaker for them is for women. This is something that even successful specialist medical doctors feel is an imperative for them.

As I note in a 2012 blog post, the decline of the African-American marriage very likely has economic causes, because as recently as 1950, when the economy was at a peak for working class men of all races (also giving rise to the "Great Migration" of blacks to the North), job prospects for women were still non-existent, and the mass incarceration phenomena had not emerged, there was no marriage gap.

70 percent of black women are unmarried, and they are more than twice as likely as white women to remain that way. Those black women who do marry are more likely than any other group of women to “marry down.” This is often chalked up to high incarceration rates—in 2009, of the nearly 1.5 million men in prison, 39 percent were black—but it’s more than that. Across all income levels, black men have dropped far behind black women professionally and educationally; women with college degrees outnumber men 2-to-1. In August, the unemployment rate among black men age 20 or older exceeded 17 percent. 

There had been no racial disparity in marriage rates in 1950, when 64% of black women were married. 

(As an aside, the dramatic change is Exhibit A in the case that Bell Curve style population genetic disparities, or "Roots" style ancient cultural legacies, aren't a good fit to explaining this change. Black women and black men have virtually the same genetic endowment, and the change in the nature of black marriage largely followed the Civil Right Movement from a status quo ante that wasn't racialized.)

The marriage gap for African-Americans was already alive and well, however, by 1965 when the Moynihan Report was released.

Another factor hidden in the statistics quoted above is that assortive marriage is on the rise. College educated women can and do depend economically on their husbands, in part, because they are more likely to be married to college educated men than they used to be, while non-college educated women can't and don't depend economically on the father's of their children, in part, because they are less likely to be married to college educated men than they used to be.

College educated white liberals, and non-Southern urban liberals generally, haven't been very concerned about "family values" because their families aren't falling apart. College educated women are married when they have children 90% of the time, and it is their falling divorce rates which have driven this trend noted by the Washington Examiner piece:

You have probably heard the statistic that half of all marriages end in divorce. This “data” point is as persistent as the notion that every piece of gum you swallow sticks to your ribs for seven years. Divorce, in fact, has been on a steady decline since the late 1970s. 

The U.S. Census Bureau measures the divorce rate by number of divorces per 1,000 married women. In 1979, that rate was 22.8. Divorce hit a 40-year low in 2015, with 16.8 divorces per 1,000 married women.

If the divorce rate were separated between the rate for college educated women and non-college educated women, the marriages of college educated women would have improved even more since the 1970s and would be significantly lower. As I noted in a blog post in 2008, quoting the Economist (via Non-Prophet):

[T]he divorce rate among college-educated women has plummeted. Of those who first tied the knot between 1975 and 1979, 29% were divorced within ten years. Among those who first married between 1990 and 1994, only 16.5% were. 

At the bottom of the education scale, the picture is reversed. Among high-school dropouts, the divorce rate rose from 38% for those who first married in 1975-79 to 46% for those who first married in 1990-94. Among those with a high school diploma but no college, it rose from 35% to 38%.

I strongly suspect that in the last decade the divorce rate divide between married women with no college and women with college degrees has grown even greater.

So, 58% of women with no college don't marry before having children, and women with no college are more than twice as likely to divorce within ten years if they do get married (38%+ v. 16.5%) as women who are college graduates. The odds of a woman with no college who has children having children while married and staying married for at least ten years are only about 25% or less versus 78% or more for women who are college graduates. 

Now, there is at least one data point which argues against my hypothesis linking marriage and economics, which I noted in a 2010 blog post:

The Pew Research Center interpreting Cenus Bureau data finds that education no longer has an impact on when a white woman will marry, or how likely she is to marry. College educated black women are more likely to marry than black women who lack a college education.

[White] women . . . ages 35 to 39, have been as likely to marry as those who did not graduate from college . . . For both groups, an analysis of 2008 figures shows 84 percent had married at some point before age 40.  

That is a significant shift from the 1990s, when young adults who didn't finish college were more likely to have wed than their better-educated counterparts, 75 percent to 69 percent. . . . 

Now, across the population, the typical age of marriage is 28 for both those who complete higher education and those who don't take that path or in some cases don't finish. . . .   

Since 1990, college-educated African-American women have been more likely to marry than their counterparts who do not have as much education . . .  That marriage gap among African-American women is sharpening[.]   

"College-educated women of all races . . . are marrying at rates similar to what the college-educated women of their mothers' generation did, but doing so later in life, and they are marrying at rates much higher than the college-educated women of their grandmothers' generation. And they have become less likely to divorce compared to their mothers' generation."

This doesn't change the fact, however, that white women who graduate from college are much more likely to have children after getting married than those who do not. As I noted in a 2013 blog post:

College-educated women typically have their first child two years after marrying. The high school graduates as a group have their first child two years before they ­marry.

And, I suspect that the higher rates of getting married at all for white women (often after they have children), doesn't change the fact that women who are college graduates are much less likely to divorce than those who are not. And, the still is a class based marriage gap, even if it is less pronounced for white women. As the Washington Examiner piece notes:

Nearly two-thirds of all adults over 25 were married in 1990. 

In 1960, 72 percent of all adults were married. Today, only half are married. Much of that shift is a delay in marriage — 18- to 25-year-olds don’t get married nearly as much as they used to. But that move away from young marriage doesn’t explain all the change. Record numbers of adults over age 25 have never been married: 23 percent of all men in 2012, compared to 9 percent in 1970, according to Pew Research. 

This drop isn’t even across classes, and it’s not a case of Wesleyan alumnae who write for HuffPo swearing off marriage as a tool of the patriarchy. The retreat from marriage is mostly a working-class thing. 

Look at adults over 25. The highly educated have long had a higher rate of marriage in this age cohort, but the gap has grown. In 1990, the college educated were about 9.5 percent more likely to married compared to those who never went to college. By 2015, the gap was 30 percent, according to Pew data.

The length of working class marriages helps to reconcile the gap between the "ever married" rate  at age 40, which is almost the same across all education levels for white women, and the percentage of adult women currently married at any given time.

And, the economic hypothesis is strengthened by the fact that unemployed men (even in "happy" marriages") are much ore likely to divorce than those who are employed, as I noted in a 2011 blog post. The study quoted there suggests that intermittent unemployment is a more important factor in working class divorce rates than stagnant wages for non-college educated men.

In a 2013 blog post I also note the greater marital stability seen in  the U.K. relative to the U.S. and consider what economic incentives could drive this trend that seems superficially contrary to the hypothesis that a stronger welfare state makes marriage weaker.

The Washington Examiner story also notes (as I have regularly observed in blog posts here) that the rise in out of wedlock births does not mean that teenage births have been on the rise. To the contrary, teenage births are on the decline and out of wedlock births now disproportionately involve adult women:

Teen motherhood was, in the 1990s, an infamous symbol of a declining morality bringing down the quality life, and spurring intergenerational poverty. In 1991 about 6 percent of all teenage girls had a baby. For black and Hispanic teens it was worse. 

By 2015 (the most recent data), the rate had dropped by more than 50 percent, falling to close to 2 percent. Every ethnic group saw a dramatic drop. Today, with the teen birth rate among blacks right around 3 percent and Hispanics below 4 percent, these demographics are now better on this score than whites were a generation ago. 

This drop in births isn’t due to abortion, or even simply to birth control. The fact — which would surprise the scolds of the 1990s — is that young people are having much less sex. Everyone is. Americans born in the 1990s (millennials and those even younger) were twice as likely as older generations to report having had no sexual partners in their early 20s. Overall, American adult sexual activity fell by 14 percent from the 1990s to the 2010s. Fewer high schoolers are having sex, according to CDC data: 41 percent in 2015, down from 48 percent in 2007.

Birth control is, however, still a very major factor in the declining teen birth rate, and in the declining overall birth rate. As the Washington Examiner piece notes:

The fertility rate in the U.S. hit its lowest level ever in 2016. Then last year, the rate dropped another 3 percent, according to the National Center for Health Statistics, down to 60.2 births per 1,000 women of childbearing age. That’s a massive one-year drop, and it meant that even though there are more women in their childbearing years than there were a decade ago, we got half a million fewer babies than in 2007.

A lot of this is due to fewer women having children a young ages, but the total number of children per lifetime per woman is also down. (Men and women with multiple partners over a lifetime, however, tend to have significantly more children than those who are married to one person for life.)

Against The Idiots Make Bad Spouses Hypothesis

Now, a very different interpretation of the class divide in out of wedlock births and divorces would be that people with less education have lower IQs on average, and that this leads them to make unwise choices regarding when to have children, to have more difficulty locating a good partner due to inferior dating prowess, and to have weaker abilities to negotiate marital disputes that threaten to get out of control.

There are good reasons to doubt this "ability" based hypothesis as an important factor, even though it may have some impact at the margins.

One is that the changes we've seen in marriage rates are recent and dramatic. If low IQ is fatal to your ability to get and stay married, and to wait to have children until you are married, why did this not happen prior to the 1970s? Why does it not happen in Japan?

One could argue that "no fault" divorce made possible something that otherwise wouldn't have happened, and this did occur at about the right time, but the divorce trends were in place before "no fault" divorce was adopted. New York State, the last state to make available "no fault" divorce experienced the same trends in non-marital child bearing and divorce that the rest of the nation did, for example. The African-American surge in non-marital child bearing similarly preceded both "no fault" divorce and mass incarceration for African-Americans.

A related notion is that as college graduation rates have greatly increased, from about 10% to about 33% from the late 1960s to the present, that the people who don't graduate from college are less competent.

This may impact marriage through intermediating economic factors. While non-college graduate men's economic prospects are mostly worse in relative terms because of changes in what the economy needs, a lot of the best and the brightest young men in the 1950s who didn't graduate from college were able to go into lucrative professions where success depended on their brains, like journalism and accounting and business management in big business, because a college credential wasn't a prerequisite to entry. Now, men seeking careers in those fields need a college degree as a ticket to entry and obtain one. Thus, the low unemployment, high wage part of the job pool historically available to non-college educated men have been shaved off, leaving the average non-college educated man a less fit provider than in the past.

But, again, it seems unlikely given all of the facts that the removal of these men from the pool of non-college educated husbands has so dramatically reduced the dating and husbanding skills of the remaining non-college educated husbands or husbands to be. The pool of non-college educated men has gone from 85% of men (college graduates used to be disproportionately male) to perhaps 60% of men (college graduates are now disproportionately women). But, the rise in divorce rates for non-college educated men and the rise in non-marital parenting for their children has surged much more quickly and not really in synch to this trend. Also, while more men are college graduates than used to be the case, far fewer men are high school dropouts than used to be the case when college was rare. So, while the cream has been skimmed, the dregs have also been removed, from an educational perspective.

Coming Attractions

I've been thinking a lot about how policy changes could lead to more stable families, because while multiple partners may arguably be better for the adults and make more economic sense in the current economy, it is pretty clear that it isn't better for the children. But, this post is already too long, and time is short, so that will have to wait for another day.