Good Economics for Hard Times Read online

  But there was an important difference afterward. Those whose houses were destroyed were given cash corresponding to the value of their houses and land, which they could use to rebuild or buy another house, or to move wherever they pleased. Forty-two percent of those whose houses were destroyed chose to move (and 27 percent of those whose houses were not destroyed moved anyway).15 Iceland is a small but well-organized country, and using tax and other records it is possible to follow the long-term economic trajectories of all the original inhabitants of the Westman Islands. Impressively, exhaustive genetic data also allows matching to their parents every descendant of those caught in the eruption.

  Using this data, researchers found that for anybody who was under twenty-five at the time of the eruption, losing a house led to large economic gains.16 By 2014, those whose parental houses were destroyed earned over $3,000 per year more than those whose parental houses were not destroyed, even though not all of them moved. The effect was concentrated on those who were young when it happened. This is partly because they were more likely to have attended college. It also seems that having to move made it more likely they found a job they were good at instead of just becoming fishermen, the one thing most people do in the Westman Islands. This would have been much easier for a young person who had not yet invested years in learning fishing. Still, people needed to be forced out (by the random munificence of the lava); those who kept their houses mostly remained, as many generations before them had, fishing and getting by.

  An even more remarkable example of this kind of inertia comes from Finland in the years just following the Second World War. As a result of fighting on the losing German side in the war, Finland was forced to cede a substantial part of its territory to the Soviet Union. The entire population of that area, some 430,000 people, 11 percent of the nation’s population, had to be evacuated and resettled in the rest of the country.17

  Before the war, the displaced population was, if anything, less urbanized and less likely to have formal employment than the rest of Finland’s people, but was otherwise very similar. Twenty-five years on, in spite of the bruises this hurried and chaotic exit must have left, the displaced population was richer than the rest, mainly because they were more likely to be mobile, urban, and formally employed. Being forced to move seemed to have loosened their moorings and made them more adventurous.

  That it takes a disaster scenario or a war to motivate people to gravitate to a location with the highest wages shows economic incentives on their own are often not sufficient to get people to move.

  DO THEY KNOW?

  Of course, one possibility is that poorer people are simply unaware of the opportunity to improve their economic situation by moving. An interesting field experiment in Bangladesh makes it clear this is not the only reason they don’t move.

  There is no legal barrier to migration within Bangladesh. Yet, even during the lean season, commonly referred to as monga (“season of hunger”) when there are very few opportunities to earn money in rural areas, few people migrate to the cities, which offer low-skill employment opportunities in construction and transportation; or even to neighboring rural areas that may have a different crop cycle. To understand why and to encourage seasonal migration, researchers decided to try out different ways of encouraging migration during monga in Rangpur in the north of Bangladesh.18 Some villagers were randomly selected by a local nongovernmental organization (NGO) to either receive information about the benefits of migration (basically what the wages were like in the cities), or the same information plus $11.50 in cash or credit (this amount was roughly the cost of travel to the city and a couple of days of food), but only if they migrated.

  The offer encouraged about a quarter (22 percent) of all households who would not have otherwise done so to send out a migrant. Most of those who migrated succeeded in finding employment. On average, those in the group who left earned about $105 during their migration, far more than they would had they stayed home. They sent or brought back $66 of that money to the families they left behind. As a result, the families who sent an extra migrant consumed on average an amazing 50 percent more calories; these families went from near starvation to a comfortable level of food consumption.

  But why did the migrants need the extra push from the NGO to decide to make the trip? Why was near starvation not enough of an impetus?

  In this case, it is very clear that information was not the binding constraint. When the NGO provided a randomly chosen group of people with information about the availability of jobs (but no incentive), the information alone had absolutely no effect. Moreover, among the people given the financial support who chose to make the trip, only around half went back during the next monga season, despite their personal experience of finding a job and making money. For these people, at least, it could not be skepticism about the job opportunities that held them back.

  In other words, despite the fact that those who do migrate, forced or otherwise, gain economically, it is hard to take seriously the idea that most people are just waiting for an occasion to give up everything and head to a richer country. Given the size of the economic rewards, there are many fewer migrants than we would expect. Something else must hold them back—we will return to this puzzle later. Before we come to that, it is useful to understand how the labor market for migrants functions, and in particular whether the gains migrants make arise at the expense of the natives, as many seem to believe.

  LIFT ALL THE BOATS?

  This question has been the object of a vigorous debate in the economics profession, but overall the evidence seems to suggest even large bouts of in-migration have very little negative impact on the wages or employment prospects of the population the immigrants join.

  The debate continues mainly because it is not usually easy to tell. Countries restrict migration, and in particular they are less likely to let people in when the economy is doing badly. Migrants also vote with their feet, and their natural tendency is to go where there are better options. For a combination of these two reasons, if you plotted the wages of nonmigrants in cities against the share of migrants in cities, you would find a nice upward-sloping line; the more migrants, the higher the wages. Good news for the pro-migration view, but perhaps entirely spurious.

  To find out the real impact of immigration on the wages of the natives, we need to look for changes in migration that are not a direct response to the wages in that city. And even that may not be enough, because both current residents and firms also vote with their feet. It could be, for example, that the influx of migrants drives out so many native workers from the city that wages do not fall for those who stay behind. If we looked only at the wages of those natives who chose to stay in the cities where migrants settled, we would entirely miss the pain of those who decided to leave. It is also possible the new migrant population attracts firms into a city at the cost of other cities, and we could miss the cost to the workers in those other cities.

  A clever attempt to get around some of these issues is David Card’s study of the Mariel boatlift.19 Between April and September of 1980, 125,000 Cubans, mostly with little or no education, arrived in Miami, after Fidel Castro unexpectedly gave a speech authorizing them to leave if they wished to. The reaction was immediate. The speech was delivered on April 20 and by the end of April people were already leaving. Many of the boatlifted settled permanently in Miami. The Miami labor force increased by 7 percent.

  What happened to wages? To find out, Card took what has come to be called a “difference in differences” approach. He compared the evolution of wages and the rate of employment of prior residents in Miami, before and after the arrival of the migrants, to the same trajectory for residents in four other “similar” cities in the United States (Atlanta, Houston, Los Angeles, and Tampa). The idea was to see if the growth in wages and jobs for all those already in Miami when the Marielitos showed up fell behind the growth in wages and jobs of comparable residents in those four other cities.

  Card found no diff
erence, either immediately after the immigrants arrived or some years later; the wages of natives were not affected by the arrival of the Marielitos. That was also true when he specifically looked at the wages of Cuban immigrants who had come over before this episode, who were probably the most similar to the new wave of Cuban arrivals and hence the most likely to be adversely affected by a new influx of immigrants.

  This study was an important step toward providing a robust answer to the question of the impact of migration. Miami was not chosen for its employment opportunities; it was just the closest landing point for the Cubans. The boatlift was unexpected, so workers and firms did not have a chance to react to it, at least in the short run (the workers by leaving, the firms by moving in). Card’s study was very influential, both for its approach and for its conclusion. It was the first to show the supply-demand model might not directly apply to immigration.

  No doubt as a result, the study was also extensively debated, with multiple rounds of rebuttals and counter-rebuttals. Perhaps no other single empirical study in economics has generated quite so much back and forth, and so much passion. A long-standing critic of the Mariel boatlift study is George Borjas, a vocal supporter of policies to shut out low-skilled migrants. Borjas reanalyzed the Mariel episode, including a larger set of cities for comparison and focusing specifically on non-Hispanic male high school dropouts, on the grounds they were the group we should be most concerned about.20 In that sample, he found that wages started going down very steeply in Miami after the boatlift arrived, compared to what was happening in the comparison cities. But a subsequent reanalysis showed these new results once again get reversed when data about Hispanic high school dropouts (who would seem to be the most obvious people to compare Cuban migrants with, but are for some reason omitted by Borjas) and women (again omitted by Borjas for no clear reason) are included.21 Moreover, studies continue to find no wage or employment effects when comparing Miami to a different set of cities where wages and employment were on very similar trends to Miami before the boatlift.22 Borjas however remains unconvinced, and the debate over the Mariel boatlift continues.23

  If you are not entirely sure of what to make of all this, you are not alone. To be blunt, it does not help that no one on either side ever changes their mind, and that opinions seem aligned with political views. Either way, it seems unreasonable to hang the future of migration policy on one episode that occurred thirty years ago in one city.

  Fortunately, inspired by Card’s work, a number of other scholars tried to identify similar episodes where migrants or refugees were sent to a place with little warning and no controls over where they should go. There is a study examining the repatriation to France of Algerians of European origin resulting from Algeria’s independence from France in 1962.24 Another study looked at the impact of massive immigration from the Soviet Union to Israel after the Soviet Union lifted the emigration restriction in 1990, which increased Israel’s population by 12 percent in a space of four years.25 Yet another looked at the impact of the large influx of European immigrants into the United States during the age of the great migration (1910–1930).26 In all of these cases, the researchers found very little adverse impact on the local population. In fact, sometimes the impacts were positive. For example, the European migrants to the United States increased overall employment in the native population, made it more likely natives would become foremen or managers, and increased industrial production.

  There is also similar evidence from the more recent influx of refugees from all over the world on the native population in Western Europe. One particularly intriguing study looks at Denmark.27 Denmark is a remarkable country in many ways, and one of them is that it keeps detailed records of each member of its population. Historically, refugees used to be sent to different cities without regard for their preferences or their ability to find a job. All that mattered was the availability of public housing and the administrative capacity to help them settle down. Between 1994 and 1998, there was a large influx of immigrants from countries as diverse as Bosnia, Afghanistan, Somalia, Iraq, Iran, Vietnam, Sri Lanka, and Lebanon, and they ended up sprinkled, more or less randomly, across Denmark. When the policy of administrative placement was abandoned in 1998, migrants most often went where their co-ethnics were already located. Therefore, the places where the first group of migrants from, say, Iraq had landed more or less by pure chance are where the new Iraqi migrants headed. As a result, some places in Denmark ended up getting a lot more migrants than others, for no good reason other than at some time between 1994 and 1998, they had spare capacity for resettlement.

  This study came to the same conclusion as the historical ones. Comparing the evolution of wages and employment of less-educated natives in cities subject to this chance influx of migrants to those in other cities, it found no evidence of negative impacts.

  Each of these studies suggests low-skilled immigrants generally do not hurt the wages and employment of the natives. But the level of rhetorical fervor in the current political debate, never mind whether it is supported by the facts, makes it hard to see past the politics of the people involved in the debate. Where, then, is there a calm, methodical voice to be found? Readers interested in the delicate art of consensus building in the economics profession may want to peruse page 267 of the (free) report on the impact of immigration edited by the US National Academy of Sciences, the most respected body for academics in the country.28 From time to time, the National Academy of Sciences convenes panels to summarize the scientific consensus on an issue. The panel for the immigration report had some fans of immigration and some immigration skeptics (including George Borjas). They had to make sure to cover the good, the bad, and the ugly, and their sentences often thread a long-winded path, but their conclusion is as close to unequivocal as you are ever going to get from a group of economists:

  “Empirical research in recent decades suggests that findings remain by and large consistent with those in The New Americans National Research Council (1997) in that, when measured over a period of more than 10 years, the impact of immigration on the wages of natives overall is very small.”

  WHAT’S SO SPECIAL ABOUT IMMIGRANTS?

  Why does the classic supply-demand theory (the more of something you have, the lower the price) not apply to immigration? It is important to get to the bottom of this question, because even if it is clearly true that low-skill wages are unaffected by immigration, unless we know why, we will always wonder if there was something special about these circumstances or the data.

  There are a number of factors that turn out to be relevant, which the basic supply-demand framework sweeps under the rug. First, the influx of a new group of workers will typically shift the demand curve to the right, which will help undo the effect of the downward slope. The newcomers spend money: they go to restaurants, they get haircuts, they go shopping. This creates jobs, and mostly jobs for other low-skilled people. As illustrated in figure 2.2, this tends to increase their wages and perhaps thus compensate for the shift in the labor supply, leaving wages and unemployment unchanged.

  FIGURE 2.2 Napkin economics redux. Why more migrants do not always lead to lower wages.

  In fact, there is evidence that if the demand channel is shut down, migration may indeed have the “expected” negative effect on natives. For a short period of time, Czech workers were allowed to work across the border in Germany. At its peak, in the border towns of Germany, up to 10 percent of the workforce was commuting from the Czech Republic. There was very little change in wages for natives when this happened, but there was a large drop in native employment because, unlike all the other episodes we discussed above, the Czechs went back home to spend their earnings. Therefore, the knock-on effects on labor demand in Germany did not happen. The immigrants may not produce growth for their new communities unless they spend their earnings there; if the money is repatriated, the economic benefits of immigration are lost to the host community.29 We will then find ourselves back in the case of figure 2.1, where
we are traveling the downward-sloping labor demand curve without a shift in labor demand to compensate.

  A second reason why low-skilled migration might push up the demand for labor is that it slows down the process of mechanization. The promise of a reliable supply of low-wage workers makes it less attractive to adopt labor-saving technologies. In December 1964, Mexican immigrant farm laborers, the braceros, were kicked out of California, precisely on the grounds that they were depressing wages for native Californians. Their exit did nothing for the natives: wages and employment did not go up.30 The reason is that as soon as the braceros were thrown out, farms in places that used to rely heavily on them did two things. First, they mechanized production. For example, for tomatoes, harvesting machines that could double the productivity per worker had existed since the 1950s, but adoption was very slow. In California, adoption rates went from near 0 percent in 1964, exactly when the braceros left, to 100 percent in 1967, while in Ohio, where there had been no braceros to speak of, adoption did not change at all during those years. Second, they switched out of the crops for which mechanization was not available. This is how California, at least temporarily, gave up on such delicacies as asparagus, fresh strawberries, lettuce, celery, and pickling cucumbers.