My latest essay concerns poverty.

the results of centrally-planned anti-poverty efforts are small, and perhaps negative (certainly very negative in the case of Communism). Decentralized capitalism, in which no one sets out to broadly reduce poverty, is the best anti-poverty program.

I wrote the essay before reading Kevin Lang’s definitive book on poverty–the one I talked about in this post. In light of his book, I might have to temper some of what I wrote.In my essay, I criticize the $20,000 poverty line, and I suggest instead using a definition of poverty that focuses on the truly destitute. But Lang offers a justification for the conventional poverty line, by positing a need to be able to participate in society.

Lang also calls into question the effectiveness of our economy at reducing poverty over the past thirty years. His point is that our economy grew, and yet the poverty rate remained stable. However, using a numerical example (I can’t help it–I live or die by numerical examples), I believe that one cannot say anything about how our economy is working to reduce poverty without knowing more about entry and exit rates into poverty.

The table below shows income dynamics for a hypothetical economy. The poverty line in this economy is somewhere between $0 and $20 (feel free to think of this as $20,000 if you prefer). Everyone earning $20 or more is above the poverty line, and everyone earning $0 is below it. The economy starts from a baseline in which there are 80 people, with an average income per person of $30 (feel free to think of this as $30,000) and a poverty rate of 25 percent. The economy then grows to an average income of $35. To reach this average level, there are three scenarios, called Case I, Case II, and Case III, in which people can move up the ladder, and the poverty rate behaves very differently in each case.

Income Levels Baseline Case I Case II Case III
$0 20 15 20 34
$20 20 20 20 20
$40 20 20 10 20
$60 20 20 20 20
$80 0 5 10 20
Average $30 $35 $35 $35
Povty rate 25% 19% 25% 30%
Exit Rate 6% 0% 17%

In the baseline year, there are 20 people each earning $0, $20, $40, and $60. With 20 people below the poverty line, the poverty rate is 25 percent.

Case I could be called “Rising tide.” Within each group, 5 people take a step up in earnings. Among the highest earners, 5 people go from $60 to $80. The $60 group stays at twenty people, because it gets joined by 5 people from the $40 group, which in turn gets joined by 5 people from the $20 group, which in turn gets joined by 5 people who previously were in poverty, earning $0. So poverty shrinks to 15 people out of 80, or roughly 19 percent. The last line of the table shows the exit rate from poverty as 6 percent. Although 25 percent of the people who were in poverty left poverty, I calculate the exit rate as the ratio of 5 leavers to 80 people in total.

Case II could be called the “Krugman scenario.” All of the increased income goes to the top two groups. 10 people move from the $60 group in the baseline to the $80 group. The $60 group is replenished by 10 people from the $40 group, but everyone else stays put. Average income goes up to $35, the poverty rate stays at 25 percent, and New York Times columnist Paul Krugman condemns the rise in inequality. The exit rate from poverty is zero.

Case III could be called the “Kling scenario.” Everyone moves up one income group from the baseline. 34 new families join the economy (either immigrants or newly-formed households), and they start out in poverty. Of course, as income keeps rising, these new families will participate in what I call The Escalation of Income.

Meanwhile, the new families keep the poverty rate high. In fact, the poverty rate actually increases to 30 percent, even though the average income has risen to $35, just as in the other cases. However, the exit rate from poverty is 17 percent, expressed as a percent of total population. In fact, everyone who was poor in the baseline year has ridden the escalator out of poverty.

The point of the exercise is that the data that Kevin Lang cites–higher average income with no reduction in the poverty rate–do not provide conclusive evidence that economic growth is no longer pulling people out of poverty. We could observe that same data even if everyone were being pulled out of poverty, depending on how many people are entering and exiting poverty.

I am not arguing that having new immigrant families and young families start out in poverty is morally satisfactory. However, if that is the reason for the stickiness of the overall poverty rate, then it changes the nature of the empirical relationship between economic growth and poverty. In particular, my claim that decentralized capitalism works great at alleviating poverty would be defensible.

But until I can prove that we are closer to the Kling scenario than to the Krugman scenario, I ought to be more cautious.