Sunday 25 October 2015

Immigration does lower native wages

There is a large and rich empirical literature dealing with the impact of immigration on native wages. Many commentators have concluded on the basis of this literature that immigration exerts little or no effect. This post asserts, to the contrary, that immigration often does in fact lower native wages. It qualifies the argument by pointing to circumstances where one would not expect immigration to have negative wage effects. 

In a simple supply-and-demand model, immigration pushes the labour supply curve to the right, leading to a decline in the equilibrium wage; competition for jobs increases, so wages get bid down. There are two main reasons why this inference could be unsound. First, labour demand might be extremely elastic. Yet this seems rather implausible to me, as I will proceed to argue. Second, immigration might simultaneously increase both labour supply and labour demand. Immigrants not only buy goods and services, thereby indirectly contributing to labour demand, but also start businesses of their own. For example, the tech giants Google, Ebay and Yahoo were all founded by immigrants. However, notwithstanding much pontificating on the subject, immigrants are not a single, homogeneous mass; some are more entrepreneurial than others. On average, high-skilled immigrants will be more likely to start businesses, while low-skilled immigrants will be more likely to compete in the labour market. All else being equal, there is some percentile of the skill distribution below which the average immigrant will increase labour supply more than labour demand. In addition, the more barriers to entry there are in an economy, the greater the advantage to incumbents, and the less likely immigrants will be to grow successful businesses. Having mentioned migrant heterogeneity then, the rest of this post will focus on low-skilled (non-entrepreneurial) immigration, arguing that it often does lower the equilibrium wage.

The literature dealing with the impact of immigration on native wages is plagued by at least three important methodological issues. First, immigrants frequently settle in areas where labour demand is high or rising. Under such circumstances, the negative effect of immigration on wages may be misidentified due to the endogeneity of labour supply and labour demand. Second, immigration into an area may cause natives to relocate elsewhere, thereby leaving total labour supply unchanged. And third, immigration may have opposing effects on wages at different points in the earnings distribution. For example, shareholders and executives may be able to pay themselves more when labour costs are lower, owing to higher profit margins. 

A seminal paper dealing with the impact of immigration on native wages is David Card’s study of the 1980 Mariel Boatlift, when Fidel Castro unexpectedly permitted more than 100,000 Cubans to emigrate to the United States over a six month period. Famously, “the Mariel influx appears to have had virtually no effect on the wages or unemployment rates of less-skilled workers, even among Cubans who had immigrated earlier.” However, George Borjas recently re-analysed the impact of the Mariel Boatlift, and found that, “The absolute wage of high school dropouts in Miami dropped dramatically, as did the wage of high school dropouts relative to that of either high school graduates or college graduates.” In another working paper, Sir Stephen Nickell and Jumana Saleheen examined the consequences of British immigration, taking care to partition their sample into different occupational groups. They observed “a significant, small, negative impact on average wages”, yet “closer examination reveals that the biggest impact is in the semi/unskilled services sector.” In 2008, the House of Lords commissioned a report on the economic impact of immigration, which garnered opinions and expertise from a number of leading researchers. According to the report, “Most of our witnesses agreed that there is some negative effect of immigration on the wages of low-skilled workers”. It does add, however, that “there were disagreements over the extent of the effects and the amount of evidence”. 

Of course, as the qualification in the House of Lords report makes clear, some scholars continue to read the econometric evidence as implying that any downward pressure on wages due to immigration is negligible. However, statistical studies in which one attempts to identify the causal impact of immigration on wages are not the only kind of evidence that can be brought to bear on the present question. There are several kinds of circumstantial evidence, which––taken alongside the statistical studies––constitute a fairly compelling case that low-skilled immigration really does depress native wages. 

The first kind of evidence comes from the effects of changes in labour supply for reasons other than immigration. Economic historians have documented a marked rise in labourers’ real wages following the scourge of the Black Death, which dramatically reduced the size of the European workforce. In fact, wages increased so rapidly that controls were implemented by a number of West European states. Though somewhat more tentative, according to several US studies, the mass-entry of women into the labour force during the second half of the 20th century depressed wages at the low end of the skill distribution, particularly among men. In both these cases, a plausibly exogenous change in labour supply influenced wages in exactly the manner predicted by the simple supply-and-demand model outlined above.

The second kind of evidence comes from public opinion. Polls consistently show that native whites in the lowest social classes are the least sanguine about immigration. Moreover, such individuals are not uniformly opposed to all categories of immigration: they are much more hostile to low-skilled immigration than to high-skilled immigration. This suggests that their opposition is at least partly motivated by economic concerns, rather than solely by attachment to national culture or xenophobic prejudice. In addition, Europeans living in EU member states are noticeably more favourable toward richer non-EU countries joining the EU (e.g., Switzerland) they are toward poorer non-EU countries joining (e.g., Macedonia). 

The third kind of evidence comes from the political activities of professionals and big businesses. If demand for labour were extremely elastic, there would be little incentive for practitioners to restrict entry into their profession through occupational licensure. In reality, a diverse array of professions have managed to get legislation passed requiring new entrants to pay extortionate fees, sit time-consuming exams, or jump through pointless bureaucratic hoops. Such measures serve to reduce the total supply of practitioners, the better to bolster the wages of incumbents. In the US, it is not merely doctors who must have a licence in order to practice, but also auctioneers, massage therapists, hair braiders, and even hypnotists. Furthermore, if demand for labour were extremely elastic, there would be little reason for big business to lobby for higher levels of immigration. Yet in reality, the Chamber of Commerce and the Confederation of British Industry––the primary big business lobbies in the US and UK respectively––consistently press their governments to let in more immigrant workers. Indeed, the infamous Koch brothers are staunchly in favour of higher immigration––a stance that has not won them many friends within the conservative ranks of the Republican Party. 

The fourth kind of evidence comes from statements made by prominent “old left” figures. Such statements constitute germane evidence insofar as the old left has conventionally seen protecting the interests of low-paid workers as its central goal; this is in contrast to the new “liberal” left, which is concerned with defending all victim groups, foreign and native. No less a luminary of the old left than Karl Marx himself recognised the negative wage effects of low-skilled immigration. In an 1869 letter, Marx wrote:
The railway to California was built by the bourgeoisie awarding itself through Congress an enormous mass of ‘public land’, that is to say, expropriating it from the workers; by importing Chinese rabble to depress wages; and finally by instituting a new off-shoot, the ‘financial aristocracy’.
And in an 1870 letter, he wrote:
Owing to the constantly increasing concentration of leaseholds, Ireland constantly sends her own surplus to the English labour market, and thus forces down wages and lowers the material and moral position of the English working class.
It is not only historical representatives of the old left who have made these sorts of statements. Bernie Sanders, the self-described socialist senator from Vermont, said the following in a recent interview with Ezra Klein: 
Open borders? No, that’s a Koch brothers proposal… that’s a right-wing proposal that essentially says there is no United States… It would make everybody in America poorer––you’re doing away with the concept of a nation state, and I don’t think there’s any country in the world that believes in that… What right-wing people in this country would love is an open-border policy. Bring in all kinds of people, work for $2 or $3 an hour, that would be great for them.”
Indeed, in a 2011 article, the conservative columnist Ed West asked, “When did opening borders to mass immigration become a ‘Left-wing’ idea?”

In conclusion, whilst immigration is capable of increasing both labour supply and labour demand, low-skilled immigration tends to increase labour supply more than labour demand, which has the effect of lowering workers’ equilibrium wage. Evidence for this mechanism can be found not just in econometric analyses of immigration, but also in changes in labour supply for reasons other than immigration, in public opinion, in the political activities of professionals and big businesses, and in statements made by prominent members of the old left.

Sunday 11 October 2015

Follow-up on homicide and gun ownership in the US

My last post inspired a number of requests for a multiple regression analysis with potential suppressor variables included. The table below shows standardised coefficients from OLS models of the state murder rate. In addition to population density (which was requested twice), Southern state, fraction in poverty, and fraction black were selected as covariates, due to their well-documented associations with the frequency of homicide in the US. These variables were all taken from the US Census Bureau


The estimate for gun ownership is quite unstable across the different models, ranging from marginally significant and positive to significant and negative. In the final model, which explains just under 70% of the variance in state murder rate, gun ownership has a negative estimate; both fraction in poverty and fraction black have positive estimates. If anyone wants to play around with the data, I'd be happy to share them. Or else they can be retrieved through the various links provided. 

Thursday 8 October 2015

Gun ownership and homicide in the US and the EU

The chart below plots estimates of gun ownership against the murder rate for EU countries (red) and US states (blue). Estimates of gun ownership for EU countries were taken from the 2007 Small Arms Survey. Estimates of gun ownership for US states were taken from a paper by Okoro et al. (2005; Pediatrics). Because Okoro et al.'s figures are from 2002 (when gun ownership was lower), and correspond to household ownership rather than simply guns per capita, I (conservatively) added 20 to each state's estimate. I say "conservatively" because this still only brings up the average gun ownership to ~60 per 100 people, whereas the Small Arms Survey estimated ownership at 89 per 100 people in 2007. Homicide figures for EU countries were taken from Eurostat, while those for the US were taken from the FBI. Figures were averaged over the years 2008-2010 to obviate sampling error in small states with few murders.


Overall, the murder rate is much higher in the US than in the EU. However, within each continent there is no clear association between gun ownership and the murder rate. Some US states are armed to the teeth (e.g., North Dakota, Idaho, Utah and Iowa), but have murder rates on a par with countries such as Ireland, Belgium and Cyprus. By contrast, other US states are not especially well-armed (e.g., New Jersey, New York, California and Maryland), but have murder rates considerably higher than nearly all EU countries. Interestingly, the four EU countries with the highest murder rates are concentrated in the North East, next to Russia.

Furthermore, Eugene Volokh at the Washington Post recently reported that there does not seem to be any association between the murder rate and the stringency of gun control laws across US states:
There’s been much talk recently — including from President Obama — about there being a substantial correlation between state-level gun death rates and state gun laws. Now correlation obviously doesn’t equal causation; there may be lots of other factors that are the true causes of both of the things that are being measured. But if we do look for now at correlation, it seems to me that the key question should focus on state total homicide rates, or perhaps (for reasons I describe below) total intentional homicide plus accidental gun death rates. And it turns out that there is essentially zero correlation between these numbers and state gun laws.
He argues, as I also would, that the appropriate measure of homicide is not murders committed with guns per capita, but simply murders per capita. The main reason being that, in the absence of guns, individuals presumably select into other methods of killing, such as knives or clubs.