

Discover more from Aporia
The Economic Illiteracy of "diversity is our strength"...
Economist Aldo Rustichini walks us through the basic economics of sensible immigration policy.
Written by Aldo Rustichini.
Mainstream pundits constantly and sententiously repeat the claim that “diversity is our strength.” And though one is tempted to call it an obvious lie, this may not be so. For diversity is indeed a strength, but only for the political class, not for the rest of the population. Diversity helps elites become and stay rich and powerful, often at the expense of others. Therefore, those who support generous immigration policies are giving a helping hand to a group that is often selfish and corrupt—and generally denounced by economically oriented leftists.
To see why this is so, we will consider a simplified, but not abstract situation. The essential facts are true. The analysis might seem complex, but the idea is simple: The political class gains by increasing the distance between rich and poor in society. It can siphon more money in a society that is poorer but more divided than in a country that is richer but homogenous. The more technical details are presented in this document.
We live in a regime with majority-rule voting and no restrictions on who can vote. The individuals in the population differ on some fundamental productive skill (e.g., intelligence, self-control, et cetera, but we’ll focus on intelligence). In a homogenous population, this trait is distributed as a normal random variable, a bell curve. We normalize it to 100 mean and standard deviation of 15.
The income of each person increases with the skill, such that the higher your skill, the higher your income. This assumption, though contradicted by some progressives, is obviously true. The evidence is overwhelming that intelligence is positively related to income (and SES). But what matters is not whether income increases with intelligence, but how fast it increases. Is the increase, for example, from 100 to 110 IQ larger or smaller than the one from 90 to 100? This is called the marginal productivity of intelligence; and if it is the same at all levels (that is, if the rate of increase is the same at all levels), then we assert that income is linear in intelligence.
Consider now a political class that wants to make the revenue from taxation as large as possible. Taxes will be collected, and revenues distributed among the population. Wealth and power for the political class depend upon how large the total revenue administered is. Suppose that what goes to politicians is a fraction of this total amount; and they want to make this slice of the pie, which is the fraction times the total amount, as large as possible.
This is obviously a simplified way to think about the interests of members of the political class. In the tradition of public choice theory, we assume that, independently of their ideological predilections, the political class becomes more powerful the larger the total amount of spending they control. This money is not necessarily pocketed, spent on booze and prostitutes, or invested in mansions. Rather, the goal is power, personal vanity, and ideological inclinations toward public projects, often those which enhance the politician’s status! The money is a means to power.
Politicians can change the composition of the populace that they govern—for example, by making the enforcement of immigration laws harder or softer. The question is: What population composition would they prefer if they are mindful of their own interests (as adduced above)?
The important policy is tax rate (which relates to the amount of public spending they control), and tax policy is decided by majority voting. The tax rate that gets more than fifty percent votes is approved. For simplicity, we’ll make this a flat tax, so we will not deal with the possibility of progressive taxation. The tax rate can be anywhere between zero percent and a ceiling of say eighty percent.
Now, let’s consider a voter with a given level of intelligence and income who is considering a tax rate proposal, let’s say forty percent. With this tax rate, he will lose income at forty percent. Everyone else will too. And the revenue will be distributed among the population. So this voter will lose his own income at forty percent and get the average income of the population also at forty percent. Suppose, for example, that his income is fifty thousand, but the average income is seventy thousand. He will get more income (in redistribution) in return for the tax rate. So the reasoning of this single voter is straightforward: He must compare his income to the average and vote for the tax rate if his income is below the average. What tax rate would we expect to prevail?
Earlier, we noted that the higher the skill (intelligence), the higher the income. The important point is how fast that increase is. Let’s consider a case with a linear benchmark. The rate of increase in income is the same for all levels of intelligence. In this case, income is also a normal random variable, a bell curve. In this simple setup, if a voter is below the average income, he or she will support taxes all the way to the ceiling. If a voter is at the mean income, then he or she is a switch voter, that is, a voter who is indifferent about taxes. He or she loses the taxed income but gains it back through redistribution.
Therefore, the fraction of the population that will vote for taxes is the fraction with income below the average. In a homogenous population, what proportion is this? Since the income is a bell curve, the proportion is fifty percent. The outcome is a tie. But if the relation between intelligence and income is non-linear, things change.
If, for example, the gain from 100 to 110 IQ is larger than 90 to 100 IQ, our predictions will change. The higher a voter’s intelligence, the higher her income, the less likely she is to vote for the tax. Consider, now, a voter who is at 100 IQ, with half of voters below her. This is the median voter. Which rate will she vote for? In this case, voters above the median IQ make much more income than those below (100-120 IQ makes relatively more than 80-100 IQ), which shifts the mean income so that the median voter is now below the average in income, and she will vote for the tax.
Of course, the political class also wants to take and keep some of this money, which means that not all the money is redistributed back into the economy. Thus, the political class cannot take too much money because they would lose votes for their proposed tax rate. The amount that they can make is larger than zero but probably smaller than they would desire. Because more money is always good.
But things change if the population changes. For example, if the political class increases the fraction of low-skilled immigrants, then that will bring down the income of the median voter; and this will increase the likelihood that he will support taxes because it will increase the gap between his income and the average income of the population. And this increases the pool of resources available to the political class. They can plunder more and still garner votes for their tax proposals.
For the political class, the optimal population is a heterogenous one with high-skill and low-skill people. Thus, there is an optimal immigration policy: No immigration makes the country too homogenous. Too much immigration might shrink the pie for a different reason.
The “sweet spot” here is easy to calculate; it depends on many parameters of the problem, such as the mean skill of the entering population (or populations) and the marginal productivity of intelligence. Typically, an intermediate value is optimal—not too large and not too small: a Goldilocks zone.
We can test the hypothesis by estimating the production function of intelligence, which is estimated from data on earnings in the UK. These data teach several interesting lessons on the way social outcomes vary when we increase the fraction of the low-skill population.
The first is that, as everyone (except blind supporters of open borders) knows, average earnings decline when this fraction increases. The second is illustrated in the figure below, which reports the payoff to the political class as a function of the fraction of the low-skill population. Looking at the two extremes, it is clear that if forced to choose between two extreme populations (either all high skill or all low skill) the political class would choose the high skill. But given the choice of a mixture of the two, the best combination is 67 percent of low-skill workers.
The anxieties surrounding the conspiracy theory often referred to as the "great replacement" might be overstated, akin to a chicken's imagined terror that the coop will be filled with foxes.
Aldo Rustichini is an Italian-born American economist. He is a professor of economics at the University of Minnesota, where he researches, among other topics, decision theory, game theory, general equilibrium theory, and bounded rationality. He has degrees in philosophy, economics, and mathematics.
Read more from Aldo: