The heritability straw man has copped another bashing, this time in the Journal of Economic Perspectives. In it, Charles Manski picks up an old line of argument by Goldberger from 1979 and argues that heritability research is uninformative for the analysis of policy.

Manski starts by arguing that heritability estimates are based on the assumption that there is no gene-environment correlation. Manski writes:

The assumption that g and e are uncorrelated is at odds with the reasonable conjecture that persons who inherit relatively strong genetic endowments tend to grow up in families with more favorable environments for child development.

Any review of discussions of heritability, whether in the peer-reviewed literature or the blogosphere, will show that his claim is generally false. The proviso that the heritability estimate is only relevant to the existing environment is usually threaded through any discussion of heritability.

It is true that gene-environment covariance can affect estimates of heritability. Yet this does not mean that existing estimates have no value, nor that there are not methods that seek to account for the covariance. For example, the use of comparisons between misdiagnosed identical twins and actual identical twins allows for bounded estimates of heritability to be developed (pdf).

Manski’s broader claim, adopted directly from Goldberger, is that even if you knew the heritability of a trait, it tells you nothing about social policy. Manski uses Goldberger’s eyeglasses example as an illustration:

Consider Goldberger’s use of distribution of eyeglasses as the intervention. For simplicity, suppose that nearsightedness derives entirely from the presence of a particular allele of a specific gene. Suppose that this gene is observable, taking the value g = 0 if a person has the allele for nearsightedness and g = 1 if he has the one that yields normal sight.

Let the outcome of interest be effective quality of sight, where “effective” means sight when augmented by eyeglasses, should they be available. A person has effective normal sight either if he has the allele for normal sight or if eyeglasses are available. A person is effectively nearsighted if that person has the allele for nearsightedness and eyeglasses are unavailable.

Now suppose that the entire population lacks eyeglasses. Then the heritability of effective quality of sight is one. What does this imply about the usefulness of distributing eyeglasses as a treatment for nearsightedness? Nothing, of course. The policy question of interest concerns effective quality of sight in a conjectured environment where eyeglasses are available. However, the available data only reveal what happens when eyeglasses are unavailable.

Manski and Goldberger may be correct that the heritability estimate is uninformative as to the efficacy of distributing eyeglasses, but it is useful in assessing other policy responses to the problem and the trade-offs between them. Is it possible to prevent the eyesight loss in the first place? Is that policy cheaper and more effective than eyeglasses? If the heritability estimate was zero, you would look to the environmental causes and ask whether the eyesight problem is more appropriately dealt with by addressing the cause rather than by distribution of eyeglasses.

There is no shortage of other areas where heritability estimates might add value. Heritability estimates can inform whether it is an effective use of resources to make sure that everyone has a university degree or is over six-foot tall. Is everyone putty in the hands of the policy maker, or are there some constraints? On a personal level, Bryan Caplan’s use of heritability in Selfish Reasons to Have More Kids is a useful input to his parenting strategy.

For me, the most salient example of the usefulness of heritability research comes from examination of the heritability of IQ among children. Among high socioeconomic status families, the heritability tends to be high. Among low socioeconomic status families, it is significantly lower. This suggests that there is significant room to improve the outcomes of the children at the bottom of the socioeconomic ladder in the early years of their life (assuming those changes have effects that persist into adulthood). Increasing heritability of IQ might be evidence that environmental disadvantages are being ameliorated and opportunity equalised.

The latter part of Manski’s paper turns to the use of genes as covariates in statistical regressions. Regression identifies statistical association and not causation, which appears to be an important point in attracting Manski to this use. Noting the wealth of data being created and the possibility of observing changes in the effect of genes as the environment changes, Manski considers that these regression exercises may assist in examining how genes and environment interact.

I don’t disagree with Manski, but at present, genome association studies have plenty of issues. First, there is the missing heritability problem. To date, the magnitude of the identified effect of genes on most traits accounts for a miniscule proportion of the trait’s heritability. This points to the important role played by heritability research to provide direction to research on genes as covariates. It also indicates that until these genes are found, heritability estimates will be more informative for social policy.

A second issue is that with 30,000 odd genes and the ability to test so many of them for correlation with traits, many are found to have a statistically significant relationship through chance. As blogged about recently by Razib, this is shown when people seek to replicate earlier results – such as when it was found that most reported genetic associations with general intelligence are probably false positives (pdf).

Finally, genome based research is now feeding back into estimates of heritability. From a recent paper:

We conducted a genome-wide analysis of 3511 unrelated adults with data on 549 692 single nucleotide polymorphisms (SNPs) and detailed phenotypes on cognitive traits. We estimate that 40% of the variation in crystallized-type intelligence and 51% of the variation in fluid-type intelligence between individuals is accounted for by linkage disequilibrium between genotyped common SNP markers and unknown causal variants. These estimates provide lower bounds for the narrow-sense heritability of the traits.

Despite all the critiques about methodology, most new studies confirm that the old “methodologically poor” heritability estimates were in the right ballpark. The problem is not that the estimates are not useful, but rather that they are not used.

Manski, C. (2011). Genes, Eyeglasses, and Social Policy Journal of Economic Perspectives, 25 (4), 83-94 DOI: 10.1257/jep.25.4.83

In my last post, I discussed Oded Galor and Omer Moav’s paper Natural Selection and the Origin of Economic Growth. As I noted then, my PhD supervisors, Juerg Weber and Boris Baer, and I have written a discussion paper that describes a simulation of the model.

In the discussion paper we consider the entry of people into the population that have a low preference for child quality – i.e. they weight child quantity more highly. Entry could be through migration or mutation. We show that if people with a low enough preference for quality enter the population, their higher fitness in the modern growth state can drive the economy back into Malthusian conditions.

To show this, we simulated a version of the model which had present at a low level in the initial population a genotype with a very low preference for educating their children (I refer to them as the strongly quantity-preferring genotype). This strongly quantity-preferring genotype has a similar fitness to other genotypes that do not educate in the Malthusian state, and declines in prevalence while the quality-preferring genotype increases.

Once the economy takes off into the modern growth state, the strongly quantity-preferring genotype has the highest fitness as it dedicates the lowest proportion of its resources to educating its children. The strongly quantity-preferring genotype increases in prevalence until, eventually, the average level of education plummets, undermining technological progress. The world returns to a Malthusian state, with high population growth eroding the income benefits of all earlier technological progress.

The following chart shows the rate of growth of population, technological progress and income per person. The first 70 to 80 generations look like the base model simulation I described in my earlier post. However, after that point, technological progress plummets to zero. For the next 150 or so generations, population growth is positive, which can occur as per person income is above subsistence. Eventually, population growth drives income down to subsistence levels.

In the next figure, you can see that the strongly quantity-preferring genotype, genotype c, grows from being a negligible part of the population to being over 90 per cent . It is this change in population composition that drives the return to Malthusian conditions (you can also see the small peak in quality-preferring types around generation 48 that kicks off the Industrial Revolution). The strongly quantity-preferring genotypes educate their children far less than the other genotypes, depressing technological progress.

There is no escape from the returned Malthusian conditions. The quality-preferring genotype will have a fitness advantage in this new Malthusian state and will increase in prevalence. Whereas that caused a take-off in economic growth the first time, this time there is no take-off. The strongly quantity-preferring types, which now dominate the population, cannot be induced to educate their children. They simply breed faster to take advantage of any technological progress spurred by the small part of the population that is educating their children.

This result could also be achieved by introducing the strongly quantity-preferring genotype into the simulation at other points in time. If it occurs after the Industrial Revolution, the timing of the return to Malthusian conditions will occur later. However, short of restricting the range of potential quality-quantity preferences, there is no way to avoid the return to Malthusian conditions in this version of model. The strongly quantity-preferring genotypes will always have a fitness advantage when income is above subsistence and their population growth will drive income back down to subsistence levels.

There are, of course, a few possible interpretations of this result. The model or assumptions may be missing an important element (or at the extreme are wrong). Humans may only have quality-quantity preferences in the growth promoting range. Or possibly, modern levels of economic growth are only transient.

As I have focussed my PhD research on the link between evolution and long-term economic growth, for months I have meant to blog on the core paper in this area, Natural Selection and the Origin of Economic Growth by Oded Galor and Omer Moav. I have held off writing this post pending finalisation some of my own related work, which I have now done.

This paper is somewhat of an outlier as I’m not aware of any other paper that models the Industrial Revolution as a result of natural selection (apart from a soon to be published paper by Galor and Michalopoulos). There is another paper by Zak and Park that examines population genetics and economic growth (a topic for another blog post) but they do not directly tackle the Industrial Revolution. In A Farewell to Alms, Greg Clark notes that Galor and Moav’s paper reignited his interest in this topic.

Galor and Moav’s paper is based on a model that has two types of people in the population. Each of these types has a different, genetically inherited preference for quality or quantity of children. The quality-preferring genotype wants their children to have higher human capital, so they invest more in their education, while the quantity-preferring genotype is more interested in raw numbers.

During the long Malthusian era in which both genotypes struggle to earn enough to subsist (i.e. during the thousands of years leading up the Industrial Revolution), the quality-preferring genotypes have a fitness advantage. As the quality-preferring genotypes are of higher quality, they earn higher wages. These higher wages are more than enough to cover education expenses, so they are also able to have more children than the quantity-preferring genotypes.

This fitness advantage leads the quality-preferring genotypes to increase in prevalence. As this occurs, technological progress increases, as the average level of education in the population drives technological progress. This in turn increases the incentive to invest in education, creating a feedback loop between technology and education.

As this goes on, the population grows. Per capita income does not increase as any technological progress is balanced out by population growth, which is the central problem of the Malthusian world.

Eventually, the rate of technological progress gets high enough to induce the quantity-preferring genotypes to invest in education. When this happens, the average level of education jumps, boosting technological progress and causing the Industrial Revolution.

During this process, the population growth rate changes. Up to the time of the Industrial Revolution, population growth increases with technological progress. However, when the level of technology leaps with the Industrial Revolution, the level of education becomes so high that population growth drops dramatically. Everyone is investing more into education than raw numbers of children.

From an evolutionary perspective, the Industrial Revolution also changes the selection pressure in the model. After the Industrial Revolution, the quality-preferring genotypes invest so much into education that they have lower fertility than the quantity-preferring genotypes. They then reduce in prevalence, their fitness advantage erased.

Galor and Moav paper work through the dynamics of the model using phase diagrams. It is not particularly easy or intuitive to see the processes working together in their paper, so my two PhD supervisors and I have just put out a discussion paper that describes simulations of the model – and shows the dynamics in a form that is easier to visually comprehend. In the chart below, you can see the dramatic jump in technological progress around generation 45 of the simulation, with per capita income growth also jumping at that time. Meanwhile, population growth drops to zero.

This second chart shows the change population composition. The quality-preferring genotype (genotype a) steadily increases in prevalence through to the Industrial Revolution, peaking at just under 5 per cent of the population. Afterwards, it is selected against.

This change in selection pressure has an interesting implication. While natural selection is the trigger of the Industrial Revolution, the population composition before and after the transition is the same. There is no difference in population composition between developed and undeveloped countries. The only time there is a difference in population composition is during the transition, when the quality-preferring genotypes peak.

In some ways, the natural selection occurring in Galor and Moav’s model is a sideshow to the main event, the quality-quantity trade-off. In a similar model by Galor and Weil, a scale effect triggered the Industrial Revolution – that is, the concept that more people leads to more ideas, so technological progress increases with population growth. I am sure that other triggers could be substituted.

That highlights the point where I am not convinced that the model is true (to the extent that a model can be). As far as human evolution relates to economic growth, I expect that inherent quality is more important (and by quality, I mean economically useful qualities) than the quality-quantity trade-off. The Industrial Revolution was possible because higher quality people were selected for in the lead-up. Higher quality people had more children as they were genetically of higher quality and they passed their high quality genes to these children. The investment in quality is choosing a high quality mate.

If quality is inherent, a high-quality person should have as many children as possible and this would have little effect on quality. For a man of low resources, his larger problem is convincing a woman to mate with him and not deciding on the right quantity-quantity mix.

The other thing that I should note is that, like most economic models, Galor and Moav’s model includes consumption with no clear evolutionary rationale (an issue I have discussed in an earlier post). Why do people in the model consume more than subsistence? If some people chose to focus all excess consumption into raising children they would come to dominate the population. This might be justified as being something to which the population has not yet adapted, but that explanation does not satisfy me.

Having made these quibbles, the model is still an impressive feat. It would not have been an easy task to create a model with technological progress, population and per capita income all following a path that resembles the last few thousand years of economic growth. There are some further issues and extensions to the model that we explore in the discussion paper I referred to above, but I’ll talk about them in my next post.

Galor, O., & Moav, O. (2002). Natural Selection and the Origin of Economic Growth The Quarterly Journal of Economics, 117 (4), 1133-1191 : 10.1162/003355302320935007

Despite being someone who is about to face the number of children question, I am not sure that I am the target audience for Caplan’s book. I don’t mean that Caplan wouldn’t recommend to me that I have more children. Rather, as someone who has thought a lot about evolution and economics and having read many of the giants on whose shoulders Caplan stands (particularly Judith Rich Harris and Julian Simon), I didn’t learn a lot from the book. As Caplan ran through the examples of twin studies showing all the different facets of a child’s personality or life outcomes that a parent has no influence over, I found myself wanting more meat and analysis. I felt similarly about his arguments for a larger population.

Having said that, and recognising that I am not the target audience, most readers would probably learn a lot. Caplan provides a fun, easy to read book that gives a great, swift overview of his case. This is the book I’ll be giving to parents, grandparents and friends who have heard me go on about twin studies and genetics. I particularly like it that Caplan gives some practicality to the swathes of findings about trait heritability.

I felt that the largest shortcoming of the book was that it does not address the third factor affecting outcomes for the child – non-shared environment. While heritability explains some of the variation in a child’s traits and outcomes, and nurture generally explains close to nothing, Caplan does not explore the research into non-shared environment. Instead, he puts the variation down to free will:

So far, researchers have failed to explain why identical twins – not to mention ordinary siblings – are so different. Discrediting popular explanations is easy, but finding credible alternatives is not. Personally, I doubt that scientists will ever account for my sons’ differences, because I think their primary source is free will. Despite genes, despite family, despite everything, human beings always have choices – and when we can make different choices, we often do.

Caplan states that several of his friends call his belief in free will his “most absurd belief”. While I don’t know all of Caplan’s beliefs, for the moment I will agree with his friends. In Judith Rich Harris’s The Nurture Assumption, she explored what this non-shared environment might be. In her case, she argued for the effect of peers. What bothered me most with Caplan’s take on free will was not that he did not agree with Harris’s suggestion, but rather, his “it’s all too hard” approach. Unlike Caplan, I expect that over the next few years we will add even further to the explanations for how non-shared environment influences children.

When Caplan came to addressing potential reasons why family size has decreased over the last 60 years, I wanted to hear his arguments in more depth. Take Caplan’s take on Gary Becker’s argument that as women now earn more, they have to give up more income to have kids:

This explanation sounds good, but it’s not as smart as it seems. Women lose more income when they take time off, but they also have a lot more income to lose. They could have worked less, earned more, and had more kids. Since men’s wages rose, too, staying home with the kids is actually more affordable for married moms than ever. If that’s too retro, women could have responded to rising wages by working more, having more kids, and using their extra riches to hire extra help.

It sounds neat, but Caplan assumes that the income effect, which would tend to increase the number of children, dominates the substitution effect, which would tend to decrease the number. It is perfectly plausible for the substitution effect to dominate and women to decide to have fewer children, but Caplan does not address this. He might be right, but as there is no depth to his discussion, it is hard to judge the strength of his argument.

Caplan does point out that in the United States, fertility bottomed out in the 1970s. This occurred despite further increases in income and Caplan uses this as evidence against any income based hypothesis. But the people having children in the 1970s are different to the people having children now. For those women who chose to have no children in the 1970s and possibly responded most strongly to the income effect, they did not contribute to the gene pool and any heritable predisposition has disappeared with them. It is the children of larger families that are having children today. Second, the net fertility rate in the United States is substantially affected by recent immigrants.

Caplan’s preferred view on the decline in fertility is that we have gained a small amount of foresight, allowing us to see the negative effects of early childhood, but not gained enough foresight to note the benefits of children when they are older. There might be some truth to this, but I expect that the other factors that Caplan dismisses are also relevant.

One point where I disagree with Caplan is around his statement that men and women see eye to eye on the number of children they wish to have. Caplan considers that this puts to bed any arguments around women having increased bargaining power. While Caplan’s statistic is true in the most basic sense, the number of children that a man or woman want are a function of a number of things. The main one of these is who the other parent will be. If a woman is paired with the man of her dreams she is likely to want more children than if she is married to a guy who showed promise but has gone nowhere. While Caplan notes that condoms have been widely available since the end of World War II, the pill gave women extra power to decide who exactly the parent is. There is some interesting scope for sexual conflict here.

When it comes to policy prescriptions arising from his position, Caplan explicitly opposes natalist policies to increase birth rates. Caplan states:

After natalists finish lamenting low birthrates, they usually get on a soapbox and demand that the government “do something about it.” There are two big reasons why I refuse to join their chorus. First, while I agree that more kids make the world a better place, I oppose social engineering – especially for such a personal decision. When people are deciding how many children to have, government ought to mind its own business.

Instead, Caplan suggests that grandparents replicate the natalist incentives privately. Given this, it is interesting that Caplan drifts into supporting natalist tax credits in his recent Cato Unbound essay (as I have commented on here). I prefer his arguments for the use of private incentives from his book than his more recent encouragement of government action.

*This is a cross-post from my blog Evolving Economics.

The heritability of a trait can vary from 0.00 to 1.00, depending on the environments from which research participants are sampled. Because we know that genes play some role in the development of any trait, the precise heritability estimate doesn’t matter in a practical sense.

Heritability depends on the amount of variability in the environmental factors that contribute to a trait. The problem is that our understanding of the factors that contribute to the development of human traits in general — and to IQ in particular — is currently so deficient that we typically do not know if the environmental factors important in the development of a particular trait are stable across testing situations, vary somewhat across those situations, or vary wildly across those situations.

In his conclusion he states:

At the very least, heritability tells us how much of the variation in IQ can be accounted for by variation in genetic factors when development occurs in an exquisitely specific range of environments. However, David S. Moore has argued that even this is not significant when we realize that the magnitude of any heritability statistic reflects the extent of variation in unidentified non-genetic factors that contribute to the development of the trait in question.

(HT: Bryan Caplan)

Through his post, Kaufman constructs a series of paper tigers, tears them down and implies that because the extreme case does not hold, we should be wary of heritability estimates. I did not find much to disagree with in his examples, but the I differed on the conclusions we should draw.

So, where I do not agree – first, the heritability estimate does matter. While I don’t think it is hugely important whether the heritability of IQ in a specific sample is 0.5 or 0.6, it is important whether the measured heritability is 0 or 0.6. As Caplan notes in his post:

My money says, for example, that the average adult IQ heritability estimate published in 2020 will exceed .5.

I think that Caplan is right (although I might have stated some conditions about the relevant sample), and Kaufman’s argument overstates how finely tuned the environment needs to be to get a meaningful heritability estimate. Heritability estimates of a sample of children growing up in extreme poverty might be much lower (or zero) but as is found again and again, once the basic requirements of a child are met, heritability estimates for IQ are consistently above 0.4. We can construct arguments that in each study there are different gene-environment interactions and so on, but if genes weren’t important in variation in IQ and the gene-environment interactions weren’t consistent to some degree, why would such consistent heritability results (and correlation between parent and child IQ) be found?

Further, these results matter. They suggest that poverty is affecting the IQ of some children, and policies could be tailored to cut this disadvantage. For children not subject to deficient environments, the high heritability of IQ should influence policies such as those for education. Children are different and the education system should take this into account.

Implicit in Kaufman’s post was the “its all too complex” argument. Social and biological sciences are complex (which is why I find them interesting). However, if we fully accepted Kaufman’s argument that “our understanding of the factors that contribute to the development of human traits … is currently so deficient that we typically do not know if the environmental factors important in the development of a particular trait are stable across testing situations”, it would put into question most of the data analysis in economics, sociology and biology. Econometrics operates on the idea of all other things being equal.

Fortunately, Kaufman has not taken the Gladwell-esque approach of suggesting that we forget about genetic factors. Kaufman suggests further research into how nature and nurture are intertwined. If it is all too complex, we should start unwinding the complexity. However, I believe that, in the meantime, this complexity does not mean that we should throw out all the results that have previously been obtained.

**This is a cross-post from my blog Evolving Economics.

As an example of the consequences of the transmission of income. Bowles and Gintis cited a paper by Hertz which showed that a son born to someone in the top decile of income had a 22.9 per cent chance of attaining that decile himself, compared to a 1.3 per cent chance for someone born to parents in the bottom decile. Conversely, a child born to parents in the top decile had only a 2.4 per cent chance of finishing in the lowest decile compared to over 31.2 per cent for those born to bottom decile parents.

As Gladwell did, Bowles and Gintis started their examination with IQ. To calculate the inheritance of income through genetically inherited IQ, Bowles and Gintis considered the correlation between parent IQ and income, the heritability of IQ from parent to child and the correlation between IQ and income for the child. Breaking this down, Bowles and Gintis used the following steps and estimates:

1. The correlation between parental income and IQ is 0.266.

2.If the parents’ genotypes are uncorrelated, the genetic correlation between the genotype of the parents and of the child is 0.5. This can be increased with assortive mating (people pairing with people more like themselves) to a maximum of one (clones mating). Bowles and Gintis use 0.6.

3.The heritability of IQ is 0.5.

4. The correlation between child income and IQ is 0.266.

Multiplying these four numbers together gives the intergenerational correlation of income due to genetically based transmission of IQ. I think there is a mistake in the calculations used by Bowles and Gintis, as they find an intergenerational correlation of 0.01, where I calculated 0.02. This leads to genetically inherited IQ variation explaining 5.3 per cent of the observed intergenerational correlation in income. Regardless of the error, this is a low proportion of the income heritability. (After I wrote this post I did a google search to find if someone had spotted this error before – and they had – on a earlier Gene Expression post on this same paper.)

I would have used some slightly higher numbers, but pushing the numbers to the edges of feasible estimates, such as increasing the correlation between income and IQ to 0.4, the genetically based correlation between parent and child IQ to 0.8 and the degree of assortive mating so that parent-child genotype correlation is 0.8 only yields an intergenerational correlation of 0.10. Genetically inherited IQ would account for approximately 26 per cent of the observed intergenerational correlation.

Unlike Gladwell, Bowles and Gintis then asked what role other genetic factors may play. By using twin studies, which provide an estimate of the degree of heritability of income (using the difference in correlation between fraternal and identical twins) and the degree of common environments of each type of twin, Bowles and Gintis estimated that genetic factors explain almost a third (0.12) of the 0.4 correlation between parent and child income. Loosening their assumptions on the degree of shared environments by identical twins compared to fraternal twins (i.e. assuming near identical environments for both identical and fraternal twins) can generate a higher estimate of the genetic basis of almost three-quarters of the variability in income.

From this, it seems that genetic inheritance plays an important role income transmission between generations. The obvious question is what these factors might be. I expect that patience or ability to delay gratification must play a role, although I would expect that there would be a broad suite of relevant personality traits. I would also expect that appearance and physical features would be relevant. Bowles and Gintis do not take their analysis to this point.

The authors finished their analysis with some consideration of other factors, and conclude that race, wealth and schooling are more important than IQ as a transmission mechanism of income across generations (although as the authors noted, they may have overestimated the importance of race by not including a measure of cognitive performance in the regression). That conclusion may be fair, but as they had already noted, there is a substantial unexplained genetic component.

This highlights the paper’s limitation, as once the specific idea that heritability of IQ is a substantial cause of intergenerational income inequality has been dented, the identification of other (but unknown) genetic factors leaves open a raft of questions about income heritability. Using Bowles and Gintis’s conservative estimates, we still have 25 per cent of income heritability being put down to genetic factors without any understanding of what these traits are and the extent of the role they play.

In their conclusion, Bowles and Gintis touch on whether policy interventions might be based on these results. They are somewhat vague in their recommendations, but suggest that rather than seeking zero intergenerational correlation, interventions should target correlations that are considered unfair. They suggest, as examples, that there are large majorities supporting compensation for inherited disabilities while intervention for good looks is not appropriate.

One thing I find interesting in an analysis of heritability such as this is that over a long enough time horizon, to the extent that someone with a trait has a fitness advantage (or disadvantage), the gene(s) behind the trait will move to fixation (or be eliminated) as long as heritability is not zero. The degree of heritability is relevant only to the rate at which this occurs and only in a short-term context. The obvious question then becomes (which is besides the point of this post) whether IQ currently yields a fitness advantage. Over a long enough time period, variation will tend to eliminate itself and Bowles and Gintis would be unable to find any evidence of IQ heritability affecting income across generations.

**This a cross-post from my blog Evolving Economics, which is my usual blogging home.

Bowles, S., & Gintis, H. (2002). The Inheritance of Inequality Journal of Economic Perspectives, 16 (3), 3-30 DOI: 10.1257/089533002760278686

End Update

Over the past few days I’ve been very disturbed…and angry. The reason is that I’ve been reading Misha Angrist and Dr. Daniel MacArthur. First, watch this video:

In the very near future you may be forced to go through a “professional” to get access to your genetic information. Professionals who will be well paid to “interpret” a complex morass of statistical data which they barely comprehend. Let’s be real here: someone who regularly reads this blog (or Dr. Daniel MacArthur or Misha’s blog) knows much more about genomics than 99% of medical doctors. And yet someone reading this blog does not have the guild certification in the eyes of the government to “appropriately” understand their own genetic information. Someone reading this blog will have to pay, either out of pocket, or through insurance, someone else for access to their own information. Let me repeat: the government and professional guilds which exist to defend the financial interests of their members are proposing that they arbitrate what you can know about your genome. A friend with a background in genomics emailed me today: “If they succeed in ramming this through, then you will not be able to access your own damn genome without a doctor standing over your shoulder.” That is my fear. Is it your fear? Do you care?

In the medium term this is all irrelevant. Sequencing will be so cheap that it will be impossible for the government and well-connected self-interested parties to prevent you from gaining access to your own genetic information. Until then, they will slow progress and the potential utility of this business. Additionally, this sector will flee the United States and go offshore, where regulatory regimes are not so strict. BGI should give glowing letters of thanks to Jeffrey Shuren and the A.M.A.! This is a power play where big organizations, the government, corporations, and professional guilds, are attempting to squelch the freedom of the consumer to further their own interests, and also strangle a nascent economic sector of start-ups as a side effect.

You are so much more than your genes. So much more than that 3 billion base pairs. But they are a start, a beginning, and how dare the government question your right to know the basic genetic building blocks of who you are. This is the same government which attempted to construct a database of genetic information on foreign leaders. We know very well then who they think should have access to this data. The Very Serious People with a great deal of Power. People with “clearance,” and “expertise,” have a right to know more about about your own DNA sequence than you do.

What can you do? What can we do? Can we affect change? I don’t know, I can’t predict the future. But this is what I’m going to do.

1) I am going to release my own 23andMe sequence into the public domain soon. I encourage everyone to download it. I would rather have someone off the street know my own genetic information than be made invisible by the government. That is my right. For now that right is not barred by law. I will exercise it.

2) Spread word of this video via social networking websites and twitter. The media needs to get the word out, but they only will if they know you care. Do you care? I hope you do. This is a power grab, this is not about safety or ethics. If it was, I assume that the “interpretative services” would be provided for free. I doubt they will be.

3) Contact your local representative in congress. I’ve never done this myself, but am going to draft a quick note. They need to be aware that people care, that this isn’t just a minor regulatory issue.

4) The online community needs to get organized. We’re not as powerful as a million doctors and a Leviathan government, but we have right on our side. They’re trying to take from us what is ours.

5) Plan B’s. We need to prepare for the worst. Which nations have the least onerous regulatory regimes? Is genomic tourism going to be necessary? How about DIYgenomics? The cost of the technology to genotype and sequence is going to crash. I know that the Los Angeles DIYbio group has a cheap cast-off sequencer. For those who can’t afford to go abroad soon we’ll be able to get access to our information in our homes. Let’s prepare for that day.

This is a call to arms, a start. I’ve been complacent about this issue, focusing more on the fascinating aspects of ancestry inference which are enabled by personal genomics. No more. I’ll be doing a lot of reading today. If you have a blog, post the video. Raise awareness. Let’s make our voices heard. If they take away our rights because we’re silent, we have only ourselves to blame. If they take aware our rights despite our efforts, we’ll set up the infrastructure for the day when we can take back what is ours.

P.S. Feel free to post info and ideas in the comments. I just literally woke up to the urgency of this issue in the past 48 hours.

Now that we are through 2010 with no crisis (that I was aware of – is this crisis still happening in private?), I thought I’d revisit Miller’s suggestion that geneticists would show more than feared about class, ethnic and race differences.

At the time I first read the article, I found it hard to characterise this information as something to fear. As Miller identifies, it would be a consequence of some interesting progress:

Once enough DNA is analysed around the world, science will have a panoramic view of human genetic variation across races, ethnicities and regions. We will start reconstructing a detailed family tree that links all living humans, discovering many surprises about mis-attributed paternity and covert mating between classes, castes, regions and ethnicities.

This sounds good to me. To understand the way genes spread as people migrated and mixed across the world will be to gain an important insight into human history.

Miller then points out that some people may be troubled when researchers start to identify genes that create physical and mental differences between populations and identify when those genes arose. Millers states:

If the shift from GWAS [genome wide association studies] to sequencing studies finds evidence of such politically awkward and morally perplexing facts, we can expect the usual range of ideological reactions, including nationalistic retro-racism from conservatives and outraged denial from blank-slate liberals.

But it is not all bad. He closes with:

The few who really understand the genetics will gain a more enlightened, live-and-let-live recognition of the biodiversity within our extraordinary species—including a clearer view of likely comparative advantages between the world’s different economies.

Reading that last sentence, the title to the article and the first paragraph appear over inflated. People will always misuse information and there will be another body of people who will make great use of it.

Looking back at Miller’s article from the vantage point of 2011, I am not sure much has changed. If anything, there has been a slow trickling of some of these ideas into spaces where they are starting to add value. GWAS studies are filling the journals and the store of population genetic data is increasing quickly. While most blank slaters continue to ignore it and the retro-racists use bits as they see fit, some of us are ploughing through it to learn something new.

Although Miller barely touches on it, the economic idea in that last sentence is interesting. If GWAS and sequencing studies result in different skills and comparative advantages being identified across the world’s populations and economies, research into economic development could be vastly changed. However, I am not convinced that we are particularly close to obtaining that sort of information. As I noted in my last post, it seems that we are some distance from taking the load of genetic information and the associated picture of human evolutionary history and being able to link it to characteristics that matter economically.

**This is a cross-post from my blog Evolving Economics.

It was only a matter of time before some economists grabbed this population genetic data and in particular, this work by L Luca Cavalli-Sforza, Paolo Menozzi and Alberto Piazza, to see whether it could shed any light on economic development. In a paper published (in one of the top economics journals) in 2009, Enrico Spolaore and Romain Wacziarg have taken data on genetic distance from the The History and Geography of Human Genes and asked whether it is correlated with differences in income between countries.

Before examining the data, Spolaore and Wacziarg proposed a model. Take an initial population that branches into two sub-populations each time period, with genetic distance between the two populations being the time since they had a common ancestor. Each sub-population has a transmitted characteristic which is represented by a number. This characteristic mutates either up or down with a 50 per cent probability each generation, so it follows a random walk. As a result, the difference in characteristics (or vertical distance) between two populations is a function of their genetic distance, with the vertical characteristics more likely to have “walked” apart as the time since the shared ancestor increases.

Next, the authors introduce technology. They assume that when a sub-population develops a new technology, other sub-populations’ ability to adopt that technology is a function of their vertical distance from the population at the technological frontier. If technology determines income, then the difference in income between two populations is the size of the relative vertical distance from the population that is at the frontier, which in turn is related to the genetic distance. The core insight from this model is that relative genetic distance and not absolute genetic distance should have a higher correlation with differences in technology.

While I am not sure this model adds much to the initial intuition, it does serve a useful purpose in that it looks to link genetic distance with income differences through differences in vertical characteristics. If genetic distance and income differences had been directly linked (positing that genetic distance is the barrier, as has been proposed within populations by Ashraf and Galor), we would not be left with the interesting question of what these characteristics are.

On the flip-side, Spolaore and Wacziarg have produced a model in which differences in vertical characteristics are a function of random drift, rather than selection. This is a touch unsatisfying, but it is hard to see how the authors could otherwise have produced the model without a theory about what those characteristics are. The model is also agnostic about how one country may develop technology as the authors assume transmitted characteristics do not have any effect on productivity. Introducing a theory of technological development could have been interesting as if certain traits make technological development more likely, there would be two effects creating the income difference – the higher probability of technological progress coupled with the barriers to diffusion.

With model in hand, Spolaore and Wacziarg turned to the population genetic data. Taking data on from 42 world populations, they matched it to countries (for which they have economic data) using information on the ethnic composition of those countries. This formed the basis of determining the genetic distance between countries. They also took a set of European population data (of 26 populations) which would allow them to do a European analysis. The regressions had to depart from the model and test the link between genetic distance and income differences directly as the data does not tell us anything about the vertical characteristics of the population.

The authors completed a mountain of regressions in analysing the data, so here are some of the headline findings. Taking the United States as the world technological frontier in 1995 (a fair assumption), the authors regressed genetic distance against the log of income and, as expected, found that income was negatively correlated with average genetic distance from the United States population. Genetic distance also had reasonably high explanatory power, accounting for 39 per cent of the variation in the sample. The chart below gives the picture. Throwing a range of other explanatory variables into the analysis such as geography and linguistic and religious differences did not materially change this result.

Spolaore and Wacziarg then ploughed deeper into the statistical analysis by creating 9,316 pairs of countries (from 137 countries) for the world sample and 325 pairs (based on 26 countries) for the European sample and assessed the link between genetic distance and income difference. When they use this broader set of pairs, as opposed to the simple comparison with the United States technological frontier, the degree of variation accounted for by genetic distance decreases, although the genetic distance still has a material effect. For example, one standard deviation change in genetic distance accounts for 16.79% of a standard deviation change in income difference when genetic distance alone is entered into the regression.

The authors also examined a range of other factors, such as Jared Diamond’s thesis about differences in geography and domesticable plants and animals. While including these factors in the analysis reduced the explanatory power of the genetic difference measure, the significance remained. The data also allowed some analysis of earlier time periods, which was in fact easier as most countries’ populations were more ethnically uniform in, say, 1500. At for the later dates, the relationship still held.

Given the agnosticism of Spolaore and Wacziarg on what the vertical characteristics driving income differences are, I hope this paper triggers some deeper examination of what is going on. What are the microeconomic mechanisms driving this result? What are the vertical characteristics that are relevant? And going the next step from Spolaore and Wacziarg’s model, how has selection affected these characteristics? Without the characteristics being subject to selection, the change in characteristics would be fairly slow. These slow changes are then hypothesised to create a substantial barrier to technological diffusion even though the populations have been separated a relatively short period. I would suggest that selection is required.

The authors suggest that more research on peaceful and non-peaceful interaction between societies may be useful to tease out the mechanisms that they have proposed. I agree that research may be interesting, but it leaves open the question which the model ignores – how did some countries get that technological lead in the first place. Do these vertical characteristics play a role in that? Asking why others did not follow does not seem as interesting as asking why some countries got the lead in the first place.

**By way of quick introduction, I am a PhD student in Western Australia and blog at Evolving Economics. I’ll be cross-posting the odd piece that might be of interest to gnxp.com readers (of which this is the first).

Finally, I know 3-D visualization is bad form, but I went for it anyway. Below is a cube which shows the positions of Gujaratis, Chinese, Mexican Americans, and Utah whites and Tuscans from the HapMap, along with a few extra samples from friends and family. Can you tell where my parents are?

This is what Dienekes also found. Aggregating various ancestral components together to be analogous to what Zack produced at K = 5, you get the bar plot below from his runs:

I assume that all ancestry analyses will find that I have a substantial minority of East Eurasian ancestry. I have a similar amount of ancestry which is obviously connected to West Eurasia. And the rest of my ancestry is going to fall into the catchall which is “South Asian,” which Reich et al. in Reconstructing Indian History argued was in fact a compound between a West Eurasian-like population (“Ancestral North Indian,” ANI) and a South Eurasian population (“Ancestral South Indian,” ASI) which was more closely related to East Eurasians than West Eurasians, though distantly so at that (modern West Eurasians are interchangeable with ANI, but ASI do not exist in unadmixed form).

Finally, here’s an analysis of chromosome 1 and its affinities to various reference populations. I’ve labelled myself. No surprises:

I am HRP002 in HAP. DOD075 in Dodecad. IN8 in BGA. I am willing to submit to any of these new grassroots ancestry projects if they want me. But I doubt I’ll find anything too surprising now. They converge upon the same rough proportions (as they should).

I’m at the stage where I want to look more deeply into the details of how long ago the “eastern” admixture occurred. It seems to come down from both parents. If it was very recent there should be some linkage disequilibrium detectable because recombination should not have broken down the allelic associations distinctive to each ethnic group yet (this is noticeable in African Americans). But I am not so sure it is recent anymore, as I’d thought. I suspect a Tibeto-Burman and Munda element were absorbed by Bengali peasants in the course of demographic expansion in what became Bangladesh between 1000 and 1500 A.D., and that ancestry is well distributed across the population now.

But even though I won’t find anything out for myself, the reason HAP and projects like it are useful is that we need better coverage of the world’s variation. There are big coarse questions which we’ve tapped out, but there are still lots of gaps to fill. I’m willing to do my part in that (or, more precisely, at this point I’ve drafted my parents into the role, since they aren’t related and so represent two independent data points for Bengal).

Addendum: I know for many people of European ancestry this sort of thing doesn’t tell them anything new. Not so for me. I always suspected East Asian admixture due to the phenotype of my extended family (and to some extent, me. I did not need to shave regularly until my 20s), but I was always curious as to its extent. Additionally, for the reasons of phenotype I had assumed my mother had very little of such ancestry while my father had a great deal. It turns out that in fact my mother may marginally be more “eastern” than my father.