Friday 29 December 2017

Assorted Thoughts #1: Applying Averages

This series contains some of the key books, studies, papers and events that I’ve read over the last year. I’ve decided to try to write one of these posts for two reasons. First, because as I finish books, I want a record of some the key insights from them both because it’s useful to refer to and because it might be interesting for other people to read. Second, because it provides a nice little glimpse into the topics that keep my occupied in a particular year and allows for me to track how my thoughts on them develop. In an attempt to publish more and shorter posts, each topic will be a separate post.

#ApplyingAverages Long Live Cohen’s D

This topic, more than any other, has occupied my mind over the course of 2017. It’s a huge shame that it’s not a topic of discussion in public discourse and I think it’s hugely important to have mainstream voices take it on. The issue flared up during the public fallout from the “Google Memo” – but it was an issue I was thinking about in the context of immigration last year. I now feel comfortable that my thoughts on this have fully formed and set them out in detail here.

Consider the following findings:

An analysis of the available FBI data by Vox's Dara Lind found that US police kill black people at disproportionate rates: Black people accounted for 31 percent of police killing victims in 2012, even though they made up just 13 percent of the US population (Vox, 2017).

--

As shown [reproductive women in the US who had] ever [had an] abortion, sterilization, and methods of contraception increase the likelihood of divorce compared to ever married women who have never used these methods of family planning from one to two times the risk of divorce, having an abortion in the past twelve months did not meet statistical significance (Fehring (2015))

--

The physiological differences between the sexes disadvantage women in strength-based and aerobic fitness tests by 20 to 40%; so for the same output women have to work harder than men. Despite the differences, there will be some women, amongst the physical elite who will achieve the entry tests for GCC roles. But these women will be more susceptible to acute short term injury than men: in the Army’s current predominantly single sex initial military training, women have a twofold higher risk of musculoskeletal (MSK) injury. The roles that require individuals to carry weight for prolonged periods are likely to be the most damaging... On recent operations women experienced a 15 to 20% higher rate of Disease Non Battle Injury (DNBI) ("Women in Ground Close Combat Review Paper", Ministry of Defence Review, December 2014).

---

Young black people are nine times more likely to be locked up in England and Wales than their white peers, according to Ministry of Justice analysis picked up by Lammy. The BAME proportion of youth prisoners rose from 25% in 2006 to 41% last year (The Guardian on the Lammy Review)

--

66% of global terrorism [in 2014 was] attributable to just four groups: Islamic State (Isis) in Iraq and Syria, Boko Haram in Nigeria, the Taliban in Afghanistan and al-Qaida (The Guardian on the Global Terrorism Index 2014)

--

Rates of estimated diagnoses of HIV infection, rates of people living with HIV, and rates of P&S syphilis were higher among MSM than among other men or women.. The rate ratios indicate disparities between MSM and other men and MSM and women. Comparing MSM to other men, the estimated rate of diagnoses of HIV infection in 2008 was 59 to 75 times as high, the estimated rate of MSM living with a diagnosis of HIV infection was 38 to 48 times as high (Table ​55), and the P&S syphilis diagnosis rate was 63 to 79 times as high (Purcell et al (2012))

--

84% of ‘grooming gang’ offenders were (South) Asian, while they only make up 7% of total UK population and that the majority of these offenders are of Pakistani origin with Muslim heritage (Qulliam (2017)).

The point here is not whether any of these findings are true, but how we should act if they were. Could it be said that, for example, on the basis that women on average “experienced a 15 to 20% higher rate of Disease Non Battle Injury (DNBI)”, they should not be allowed in to serve in direct combat roles? Should the different infection rates among homosexuals vs  heterosexuals mean they shouldn’t be able to donate blood, or, as Heritage stated back in 1994, the “risk of AIDS is itself sufficient reason to deny gays the privilege of serving in the U.S. military”? Or that given we see average differences between races in the crime stats that the criminal justice system is prejudiced? Or that given that Muslim groups are responsible for a majority of global terrorism that we should have a ban on Muslims entering the country?

You can clearly see why I think this area requires mainstream voices actively talking about the issue. These are the implications that reactionaries and nativists come to. Without an intellectually (and, more particularly) statistically response to these policies, a raft of illiberal policies or policies that can cast doubt on our institutions can be justified. And without actually engaging with the statistics or the arguments, we are potentially left vulnerable to believing dead dogmas, or worse, having our institutions affected.

The answer to the crude nativist applying averages, I believe, is that we have to acknowledge average differences and emphasise that we are acknowledging the average not the whole. Sam wrote about how and why it was important to do this:

Think about the claim that “Women generally have a stronger interest in people rather than things, relative to men.”… For a given man and a given woman who seem similar in other respects, this claim sounds like it’s saying that each man will be more “interested in things” than each woman.

But it doesn’t! This is what my friend “the Anonymous Mugwump” refers to as a difficulty in “applying averages”. We tend to take claims about groups as claims about each individual within those groups. That means that claims about men or women being more inclined towards one thing than another sound both obviously false and easily rejected, and very insulting.

For every one hundred people, there might be ten men and nine women who are equally “interested in things”. This would mean that on average men are substantially more likely to be “interested in things” than women, but that as individuals none of those women are any less interested than those men and that there are millions of women who are just as “interested in things” as any man is. Applying the average to the individuals would be a very silly (if easily made) mistake.

Sam goes onto say:

Group differences, in other words, tell us next to nothing about the traits of a given man or woman within those groups — there are lots of neurotic men and lots of genius women. Very simple information about any individual is going to tell you much much more than whatever the distribution of attributes is for a group they happen to be part of.

I want to elaborate on Sam’s points here. Ben gave a very good talk at the Adam Smith Institute’s Forum in 2016 about how granular information can help us deal with statistical discrimination. For example, people may use average differences in conviction rates between races as information for hiring practices when they don’t have further information. The way to deal with this is to provide information about the individual: did they actually go to prison? Do they have the qualifications? When granular, individuated information is used, it makes using crude averages redundant. Take the example that Ben always uses from Air BnB reviews, as per Cui et al (2017):

We find that when guest accounts have no review, the average acceptance rate of White guests is 48%, and the average acceptance rate of African American guests is 29%. In other words, guests with White names are accepted 19 percentage points more often than those with African American names (p-value =0:0002)… When there is one positive review, the acceptance rate of White guests is 56%, and the acceptance rate of African American guests is 58%. Note that, irrespective of guests' race, the acceptance rate rises when a guest's quality is validated by a positive review. In this case, the acceptance rates between White and African American guests are statistically indistinguishable (p-value =0:8774).That is, discrimination is eliminated in the presence of one positive review.

And it’s not just new information which can be identified and utilised but also that technology can develop so that we can. The UK government justified its ban on gay men donating blood on the basis of the above mentioned discrepancy between straight and gay men. This policy was rightly reversed following a determination that “new testing systems were accurate and donors were good at complying with the rules.” Technology and free markets, operating not with "optimal" but more and more information, are key to fighting the use of crude averages.

There is a further reason why applying averages in a crude way is bad: it is encourages intellectually laziness and denies the complexity of the world around us. It purports to end the debate with the use of average statistics but issues and policy conclusions are almost always more complex. It is simple to say, for example, that as a result of the disparity in those subject to Stop and Search powers that the police are racist. But what if there are other considerations which do not rely on maligning an entire police force? And, for your information, there is suggestive evidence this is true, as per Miller (2000):

The research shows resident population measures are very different from populations actually available to be stopped and searched. Specifically, the research suggests that available populations tend to include larger proportions of people from minority ethnic backgrounds than resident populations. Furthermore, when statistics on stops and searches are compared with available populations, they do not show any general pattern of bias against those from minority ethnic backgrounds, although there are some specific exceptions… This suggest that stops and searches are generally targeted at areas where there are crime problems.

The same could be said about police murdering African Americans. It’s so easy to refer average differences between shooting White and Black Americans and conclude that the police in the U.S must be racist and gunning down black people – but what if is something else? I think there is tentative evidence to suggest that its related to violence used in particular areas, rather than racist motivations (Klinger et al (2015)), the rates do not appear to be disproportionate to the levels of average crime (MacDonald (2016/7)) and experimental evidence suggests that  police officers are more delayed in using force against Black people because of a fear of social reprisal (James et al (2016)).

These studies are not conclusive and there are conflicting ones, but the point is that the debate cannot end with the use of crude averages. Applying averages in such a ham fisted way (e.g. ban Muslims because of average differences in committing terror) stops us from properly considering controls and mitigation measures. I want to explore two issues in a bit more detail to tease out some nuances.

First, consider Donald Trump’s “Muslim ban” on 7 Muslim majority nations. The argument is easy to understand: Muslims are committing terrorism, so let’s stop Muslim immigration. To avoid being intellectually barren, we need to provide a response that does not deny the (global) disproportionality of self-described Muslims being involved in terror activities. Here are some arguments – not fully made - that I think weigh in favour of not having a blanket ban:

1. Refugees are already extensively vetted, so there’s no reason to ban a whole group of people in the first place (see a form of this argument from Natasha Hall in the Washington Post).

Response: Yes, but what about the desecendants of those refugees or migrants who come here and then commit an act of terror?

2. Terrorism is incredibly rare and, in particular, the terrorism which comes from refugees is extremely rare, as per Nowrasteh (2016):

Including those murdered in the terrorist attacks of September 11, 2001 (9/11), the chance of an American perishing in a terrorist attack on U.S. soil that was committed by a foreigner over the 41-year period studied here is 1 in 3.6 million per year. The hazard posed by foreigners who entered on different visa categories varies considerably. For instance, the chance of an American being murdered in a terrorist attack caused by a refugee is 1 in 3.64 billion per year while the chance of being murdered in an attack committed by an illegal immigrant is an astronomical 1 in 10.9 billion per year.

Response: even though its rare, terrorism poses a unique quasi-existential threat to Western civilisation, see this post of mine for an elaboration of this argument.

Response to the response: yes, but even considering that threat, the utility gained from providing new homes to refugees or immigrants outweighs this downside. Evidence points:

(a) There are major economic benefits to migrants who come here accompanies with marginal or neutral outcomes on the economic outcomes of natives (see The Empirics of The Places We Go).

(b) The amount of Muslims who are terrorists globally represents 0.0066% of the Muslim population (source), suggesting that the amount of people who can have utility gains is high. To be more particular, there a Muslim population in the UK of approximately 2.7million Muslims. There are, according to MI5, approximately 13 attacks stopped between 2013 and 2017. There are, further, 20,000 people who are considered to be at risk of being involved in terrorist activity, a significant minority of that 2.7million figure.

3. It is better that, if there are terrorists and criminals amongst migrants, that they be in a country with high state capacity and the rule of law where they are likely to be caught than their home countries where they are likely to carry out crimes with impunity. This has an empirical basis: 1 standard deviation increase in migration is associated with 1/3 of a standard deviation decrease in civil conflict in the origin state (see Preotu (2016)).

Response: But you’re increasing the crime rate that natives will be subject to! Further, or in the alternative, the purpose of the nation state is to protect its own populace before others!

Response to the response: I’m not sure that the nationalist philosophies justify anything more than deriving some sense of enrichment from your own culture. The interest that we derive from our own cultures seem incapable to me of justifying any civil or political obligations on others. And even if they did, I think they would be part of the equation in coming to a policy with all of the trade-offs including, as described above, the massive utility gains for the migrants themselves. 

In any event, if we look at the data from Germany suggests that we are not seeing levels of crime that outweigh the utility given to migrants themselves. In particular, Gehrsitz and Ungerer (2017) find that with respect to crime rates, they find “at best muted increases in criminal activity”. In particular, they found a whole one standard deviation increase in migrant flow is associated with about 95 additional crimes per 100,000 people. It’s also worth noting that this aligns with official data:

… the statistics compiled by the authorities also show that the probability is no higher among refugees than in the domestic population. According to police crime statistics, the number of criminal acts increased by about 4 percent in 2015 over the previous year. The increase was mainly attributable to a rise in asylum- and visa-related offences. If these offences are factored out of the equation, the number of criminal acts remained virtually constant, even though the number of people in the country had increased by hundreds of thousands (De Spigel (2017)).

I’m hoping the above back and forth – which I hope others and I will expand on in the future – shows why the use of crude statistics is particularly unthinking. The second example is the decision to allow women into front line combat roles in the UK. If you look at the one statistic quoted above in isolation, you see that there may be a case for not allowing women into direct combat roles. But the assessment is a perfect example in showing how policy considerations are more complex than members of the alt right who latch onto such statistics would have:

The review studied 21 factors that contribute to CE [combat effectiveness], of which physiology and team cohesion are the most relevant; these were considered under separate workstrands. The review assessed that one of the factors will be improved by the inclusion of women, seven are neutral or multi directional, eleven are likely to have a negative impact on CE and in two the impact was unknown.

The review goes onto note that with mitigation measures, these 11 negative factors drop to 3. The outstanding three negative measures are difficult to mitigate against and will be kept onto review. If we followed a crude way of applying averages, we would not have tried to identify these mitigation measures in the form of minor changes, increased resilience training etc. etc.
To summarise: we shouldn’t use the average as the whole, but nor should we deny the facts about average differences. The reason why we apply averages in such a crude way is because firstly, its not as illuminating as noting the degree of overlap between group differences; it leads to lazy thinking on both the left and right, ignoring that there are complex considerations and controls that mitigate against their conclusions.

But I want to be clear about something I am not saying. Am I saying that averages differences should never be used? No. I am saying that where we have individualised, granular information or we have the capability of obtaining it, we shouldn’t use average differences for the purposes of making policies that affect individuals. In addition, where using an average difference is completely out of line with the proportions involved, we should air on the side of not using the average: for example, the Green Team may be 50% more likely to kill, but if that amounts to a 2% chance that any member of the Green Team may kill, we cannot use a sledge hammer to crack the nut that is the killers of the Green Team.

And this brings me to the final point: why should we try to avoid using averages aside from the fact they may not illuminate as much as statistics about variations within groups and it makes people unthinking and lazy? I think one potential argument is that it reinforces a world view not based on treating people as individuals. Individualism matters: it is associated with good outcomes like a nations polity score (Gorodnichenko and Roland (2015) which in turn is related to economic outcomes (Kyriacou (2015)).

Methodological individualism is important because it tells us more about the world. What we mean by methodological individualism is that the unit of social and moral life for causal and moral purposes is the individual. An ideology or policy preference based on using averages as a unit is at odds with the correct way of scientific and historical analysis a la Karl Popper. He alludes to just as much in the Poverty of Historicism: “Unable to ascertain what is the minds of so many individuals, he [the historicist] must try to simplify his problems by eliminating individual differences” (p.90).  

When we, as a matter of policy, start treating averages as the basis for policy, rather than individuals (even when that is informed by an understanding of group differences), we end up undermining the moral and other status of individuals. I consider it to be no surprise that hard right individuals, nativists and extreme leftists utilise collectives and averages at the expense of individuals – it is the core of their ideology. It explains why so much of what they do is at odds with liberal principles of due process and the rule of law. 

I appreciate that some of the above appears to be abstract and may, to some, appear to be obvious. But I am always struck at how selective people’s approach to applying averages really is. I wouldn’t try to make a claim about how wide spread such inconsistencies are – there is no data – but I would ask people to apply, consistently, the need to for controls, the need to appreciate overlapping bell curves, the need to assess trade-offs, the need to avoid reinforcing a norm contrary to a fundamental Western norm: individualism. The next post is about the need to reinforce norms during the Trump administration.  

Sunday 18 December 2016

Mein weltanschauung

My friends Ben and Sam have both written posts about which articles and books have influenced their thinking. There’s not a whole world of difference between us in terms of our core beliefs but I thought I would write a similar list. This is pure indulgence on my part, so if you're not interested in my views, this will bring you no enjoyment. In addition, because I am incapable of writing anything short, I’m going to provide a bit of an explanation about my choices and how it interacts with my world view. I have not fully elaborated or  provided all the sources for my views - this post is supposed to be a list of book that have influenced me, rather than a full bibliography of everything that supports the views below and I don't want to take the piss much too much. 

Whiggishness 


Better Angels of Our Nature - Stephen Pinker / The Rational Optimist by Matt Ridley / The Great Escape by Angus Deaton 

These three books are responsible for my views about the general progress of the world. There have been great rises in wealth, the decline in poverty, the rise of liberal democratic political systems, the historic declines in crime and war, and the incredible achievements on a range of social issues (sexual rights, womens’ rights, gay rights, and now, hopefully, transrights). It’s the source of “Mugwump’s Law” (not a thing) which is that they more you want to change the status quo, the more your views about the status quo are probably wrong.

In a similar vein, The End of History (which I’ll return to below) is part of this belief. The progress we have seen is, at the very least, linked to the economic and political systems we have created in the West and are now proliferating slowly in the world - but is also part of the reason I moved away from being a proponent of democratic intervention. I previously believed, in principle, with the idea of removing authoritarian governments because they were authoritarian. But then I changed my mind - as Mueller (2014) points out in his essay “Did History End? Assessing the Fukuyama Thesis”: 

The central policy implication of the experience with the remarkable rise of democracy and capitalism is to suggest that, if trends are on one’s side, it may well be best not to work too strenuously to move them along... efforts to impose them are likely to be unnecessary and can be costly and even counterproductive.... People do not seem to need a lot of persuasion to find appeal in shining cities on hills that are stable, productive, and open even if some of the luster wears off as they get closer.

Incidentally, the other main reason for democratic intervention losing favour with me was the work of Scott Walker. In his 2014 study (“Does Forced Democratisation Work?”) he looks at two types of U.S. democratic interventions: (1) intervention followed by push for democracy (Variant 1) and (2) interventions in autocratic states (Variant 2). In respect of the latter, he finds that 8 of the 14 countries that fit his category, meet the definition of democracy (i.e., a success rate of just over 50%) which isn't too bad. However, the results exclude Afghanistan and Iraq, and the results don't hold when you look at a different measures (Freedom House Scores). Looking at interventions in autocracies (Variant 2), he finds that only approximately 25% of interventions are successful when looking at the Polity IV score. 

However, my view is that the aim is not for these countries to become Norway in the short or medium term, the aim is to improve the general score of polities. Here, the results are encouraging. Looking at Variant 2 countries, the average Polity VI score increases from 0.56 in 1989 to 1.11 in 2009.  But the general finding is enough to put me off the principle of democratic intervention; I continue to agree with attempts at democratisation following our security and/or humanitarian interventions. 


Institutions

But I’ve jumped the gun in starting with my whiggish views. I believe in the importance of liberal democracy - a belief that comes from more studies and books than I can possibly name. So here’s a few particular views and their sources.

Economic Related

The Long Divergence: How Islamic Law Held Back the Middle East - Timur Kuran / Why Nations Fail - Acemoglu and Robinson / Fragile  by Design  - Calomoris and Haber — Economic Origins of Roman Christianity - Robert Ekelund and Robert Tollison 

I was very late to reading about new institutional economics and my primer wasn't Douglas North, but Timur Kuran. His book, which I have provided an overview of here, provides an institutional explanation for why the Middle East became economically retarded. He places importance on the form of contracts, legal institutions like corporations, the (lack of) importance of Islamic restrictions on interest and how all of the aforementioned links with Islamic rules about apostasy, dhimmis and local court enforcement. 

This led me onto read more new institutional economic books. One of the greatest differences between me and my ‘Twitter crowd’ is that I believe that democracies provide better outcomes. I have explained the enormous weight of the evidence about how democracies are better are protecting fundamental rights. It is a common refrain that respect for liberal rights and the rule of law are responsible for good outcomes, not democracies. I will explain why I disagree with that view at some point in the future, but new institutional economic provides me with comfort in the view that democracies have better economic outcomes. 

Fragile by Design: The Political Origins of Banking Crises and Scarce Credit, a fairly convincing book written by Calomiris and Haber provides one example of why this is the case: democracies generally have more stable banking systems that provide broad credit. This is through a number of nterlinked mechanisms: an autocracy cannot give adequate guarantees that it will not expropriate the assets of bank investors, minority shareholders and debtors. Accordingly, they must ‘compensate the banks investors and depositors for accepting the risk that the banks assets may be expropriated’ (p.33). They do this by providing the banks with rents, commonly in the form of a monopoly right or privileges (p.44).  

Banks like these – effectively nationalised or, at the least dependent on the teat of the autocrat  – provide scarce credit (because most of the money goes to the autocrat himself and/or because they have monopoly rights, interest rates are high). They are also unstable: both the autocrat and the bank insider’s discretion misallocate and squander resources (by investing in either in the autocrat or the insider’s own businesses). These are all self-reinforcing mechanisms: because credit is not widely distributed, there is no need for the institutions that enforce contracts (this includes the rule of law which is required to independently adjudicate and enforce contracts).

Democracies, by contrast, protect against these risks more consistently than autocratic governments. In much the same fashion, Why Nations Fail (which is, unlike Fragile by Design, a moderately problematic book), provides the same mechanisms for general economic development. The requirement of autocrats to stay in power leads to extractive institutions; inefficient institutions that arise because the autocrats needs to provide rents and privileges to strongmen - to forestall any risk that they take over, and also to provide them with funds. The risk of economic actors becoming powerful is an additional reason why autocrats create a network of inefficiencies. 

Just a brief little caveat: I do not believe in pure democratic systems. I believe in liberal democracies: key checks and balances that try to forestall the bad elements of democracy. I think there should be stringent constitutions, courts with the power to strike down laws, an unelected upper chamber, party political leaders should have the ability to control candidates and no matter should be put before the people in a referendum. The electoral failures of Brexit and Trump are failures of gate-keeping and a rejection of trustee models of representation.
The literature on the importance of institutions above has also led to views I have about the political decay  that autocracy causes. I have written about how the origins of authoritarianism in the Middle East are internal, domestic institutions (see this post for a collation of this view). This view contributes to my view, explained, below about disagreeing with democratic intervention: liberal democracies rely on a whole host of institutional developments over centuries, the most important of which is civil society. 

Political Structure Related

Why does government fail so often? - Peter Schuck / The Silent Epidemic: Coal and the Hidden Threat to Health - Lockwood / Capitalism and Freedom - Milton Friedman / On Liberty - John Stuart Mill 

I am not an anarchist and I am not a libertarian. I believe in the power of the state to carry out certain services, to fund armies, to provide the funds for certain things (such as healthcare or the universal basic income). But, I have a huge scepticism of using the state. The high threshold for governmental action has it sources in much of my economic views. The book that started me on my path to being an economic liberal was Milton Friedman’s Capitalism and Freedom. If I read the book now, I’m sure I’d nitpick and not buy everything but it remains the case that as a 17 year old it had a massive impact on me. The same for On Liberty - I was an ardent defender of Mill throughout my university years: it gave me a respect for limited government that I have yet to shake off. 

This suspicion of government power has, like many of my views, been nuanced and tangled by an array of subsequent studies, but the fundamental point remains: the more limited the scope of government, usually the better. In Why Do Governments Fail So Often? Schuck writes about a study by Winston (2006) which looked at pretty much every single study that assessed U.S. federal government programmes that tried to handle ‘market failures’ (in particular, market power, information inequalities and externalities and public goods). He summarises the conclusions:
Thirty years of empirical evidence on the efficacy of market failure policies initiated primarily by the federal government but also by the states, suggests that the welfare cost of government failure may be considerably greater than that of market failure... Some policies have forced the U.S. economy to incur costs in situations where no serious market failure exists, while others, in situations where costly market failures do exist, could have improved resource allocation in a much more efficient manner  
... my assessment of the empirical evidence reveals a surprising degree of consensus about the paucity of major policy successes in correcting a market failure efficiently.. Generally, my fundamental conclusions are not influenced by studies that use a particular methodology. In fact, researchers who used vastly different techniques to assess specific policies often reached very similar conclusions (p.52, p.23, p.63). 
Schuck carries out a similar exercise in his book. He looks at assessments from Government Accountability Office, the Office of Management and Budget and several leading think tanks from across the ideological spectrum. His results are in line with Winston’s:  
We found more than 270 such assessments, some of which will be cited in later chapters. Only a small number of these assessments could be considered positive; the vast majority were either clearly negative or showed mixed results (p.23).
But I’m not a maniac. To briefly explain my ideological evolution: I was a social democrat, I became an anarcho-capitalist, then a minarchist - and since then I’ve made the trip to where I am today: a neoliberal (see Sam’s definition) with libertarian tendencies. The reason I’ve moved back toward the centre is, predominantly, because of the literature surrounding my foreign policy views and the environment. I’ll be talking about the former below, but I want to briefly cover the main book that made me change my mind about being a minarchist: The Silent Epidemic: Coal and the Hidden Threat to Health. In the margin of the following extract, which talks about the effect of the Clean Air Act, I have a hand written note which reads: “we need this!”: 

The improvements in air quality were thought to be primarily due to reductions in particulate matter and ozone. In this retrospective analysis, the modeling predicted an annual reduction of 184,000 premature deaths, 674 cases of chronic bronchitis, over 22 million lost days at work, and other outcomes...  
The EPA concluded that the total monetized health benefits from the Act during the 20-year period ranged between $5.6 and $49.4 trillion. The central estimate for benefits was $22.2 trillion. During that period, the costs to comply with the act were estimated to be approximately $0.5 trillion. Thus the net direct benefits were between $5.1 and $48.9 trillion, with a central estimate of $21.7 trillion...  
By the year 2020 the scenario predicted by the amended Act avoids 230,000 premature deaths among adults age 30 and above each year. The model also predicts avoiding the deaths of 280 infants each year. The monetary value of these two causes was set at $1.7 trillion for adults and $2.5 billion for infants (p.191-199). 

The Role of Ideas

Ideological Origins of the American Revolution - Bernard Baliyn / Roads to the Temple - Leon Aron / End of History - Francis Fukuyama

The central role of ideas plays a massive part of my world view. I know its cliche but ideas matter - but the reason why its not so cliche is the tendency for people to give primacy to structural issues like poverty, inequality or culture in explaining historically important events. I have published a review of Roads to the Temple (which you can read here) and explained the implications of this in The Empirics of Free Speech, Part I

To move on to the causal link between the progress we’ve made in terms of liberal democracy and free speech, Mueller is probably right when he says that 
Democracy’s rise has, it seems, essentially been the result of a 200‐year competition of ideas, rather than the necessary or incidental consequence of grander changes in social, cultural, economic, or historic patterns. It has triumphed because the idea that democracy is a superior form of government, ably executed and skilfully promoted—or marketed—at one point in the world’s history, has increasingly managed to catch on.  
By way of an elaboration I want to briefly talk about two books. The first is Leon Aron’s Roads to a Temple: Truth, Memory, Ideas and Ideals in the Making of the Russian Revolution, 1987-1991. I have reviewed this book elsewhere but the core idea is that freedom of speech allowed ‘every institution – political economic and social – to be subjected to trial by truth and conscience’ (p.51). It is following this process of self-discovery and criticism that surveys showed ‘solid majorities favour some key features of liberal capitalism’ (p.32-3). It then when people accept ‘alternatives to the current view’ does a ‘pre-revolutionary situation... become a revolutionary crises’ (p.20).  
The second is Bernard Bailyn’s Ideological Origins of the American Revolution which focuses on the pamphlets that were ‘the literature of the Revolution’ and ‘convey[ed] scorn, anger,  and indignation’ that were ‘probings, speculations, theories’ which ‘were not mere mental gymnastics’ but were they provided the ‘grounds of resistance.’ (p.8, p.18, p.231). Bailyn shows that the American Revolution was    
... above all an ideological, constitutional, political struggle... [and] intellectual developments in the decade before Independence led to radical idealisation and conceptualisation of the previous century and half of American experience, and that it was this intimate relationship between Revolutionary thought [as expressed in the pamphlets] and the circumstances of life in the eighteenth-century that endowed the Revolution with its peculiar force...  (p.x-xi).   
These two books are not quirky artefacts of the historical record.[1] 
[1] Fukuyama (1992) provides a more general history which shows the roles of ideas:  
‘The critical weakness that eventually toppled these strong states was in the last analysis a failure of legitimacy—that is, a crisis on the level of ideas. Legitimacy is not justice or right in an absolute sense; it is a relative concept that exists in people's subjective perceptions. All regimes capable of effective action must be based on some principle of legitimacy. A tyrant can rule his children, old men, or perhaps his wife by force, if he is physically stronger than they are, but he is not likely to be able to rule more than two or three people in this fashion and certainly not a nation of millions... It is clearly not the case that a regime needs to establish legitimate authority for the greater part of its population in order to survive. There are numerous contemporary examples of minority dictatorships that are actively hated by large parts of their populations, but have succeeded in staying in power for decades.... When we speak of a crisis of legitimacy in an authoritarian system, then, we speak of a crisis within those elites whose cohesion is essential for the regime to act effectively’ (p.15-16). 
I think this emphasis on the role of elites to the detriment of force is misplaced but that doesn’t change his core argument that, in the spirit of both Aron and Bailyn, ‘there was a remarkable consistency in the democratic transitions in Southern Europe, Latin America... [With a couple of exceptions] there was not one single instance in which the old regime was forced from power through violent upheaval or revolution... [rather] it was ultimately made possible by a growing belief that democracy was the only legitimate source of authority in the modern world’ (p.21).

This belief in the role of ideas is why I don't think blowback is a thing when it comes to explaining terrorism. But, in the same vein, the lack of specificity from Eustonites when they talk about ideological explanations is frustrating. The more thoughtful people don't talk about the role of Islam in terrorism, but the role of Islamism. But, for reasons I’ve laid out before, I don’t find this persuasive either. The common refrain of people like Majad Nawaz is “if you believe stoning is right, how can Islamism not be linked to violence?” (or something like that). One of the problems with this view is that they then conflate it with critical views about democracy, womens’ rights and views about free speech. 

If we’re more specific about ideologies, this criticism falls away. And this brings me to a paper from Fair et al (2012) which looks at the predictive power of Islamism as opposed to “Jihadism” - the results are unsurprising:
…neither religious practice nor support for political Islam is related to support for militant groups. However, Pakistanis who believe jihad is both an external militarized struggle and that it can be waged by individuals are more supportive of violent groups than those who believe it is an internal struggle for righteousness.

I talk about this study and half a dozens showing why Eustonites are wrong in this post. Finally, There’s one other extract which entwines my views about progress and the role of ideas - its from Zack Beauchamp’s review of The Great Escape published in Think Progress. He writes: 

it suggests that there’s a third type of innovation, beyond those of science and business, that propels humanity forward: moral advancement… Once invented, moral advancements can’t be contained inside national borders — another similarity between them and Deaton’s technologies. A belief in the fundamental moral equality of persons and the attendant democratic institutions has spread globally. Democracy is the world’s dominant form of government and belief in human rights is increasingly transcending national borders. 
Once invented, moral advancements can’t be contained inside national borders — another similarity between them and Deaton’s technologies. A belief in the fundamental moral equality of persons and the attendant democratic institutions has spread globally. Democracy is the world’s dominant form of government and belief in human rights is increasingly transcending national borders.

Liberal Nationalism

The Limits of Nationalism - Chaim Gans / Just Zionism - Chaim Gans / Liberal Nationalism - Yael Tamir / On Nationality - David Miller / Politics in the Vernacular - Will Kymlica / Multicultural Citizenship - Will Kymlica 

I consider myself a liberal nationalist - which probably comes as a surprise to some people. The refrain of a nativist that I and people who share my view are ‘rootless cosmopolitans’ has always made me laugh: I love my country and consider it to be a fundamental part of my identity. I have very complicated views about liberal nationalism that I’ll be describing in Part III of my immigration series so for now I’ll just give a few relevant extracts from the above books.

Kymlicka in Multicultural Citizenship: A Liberal Theory of Minority Rights
…freedom involves making choice amongst various options, and our societal culture not only provides these options, but also makes them meaningful to us. ...societal cultures involve 'a shared vocabulary of tradition and convention' which underlies a full range of social practices and institutions.... To understand the meaning of a social practice, therefore, requires understanding this 'shared vocabulary’ – that, is, understanding the language and history which constitute that vocabulary.   
Whether or not a course of action has any significance for us depends on whether, and how, our language renders vivid to us the point of that activity. And the way in which our language renders vivid these activities is shaped by our history, our ‘traditions and conventions’. Understanding these cultural narratives is a precondition of making intelligent judgment about how to lead our lives. In this sense, our culture not only provides options, it also provides the spectacles through which we identify experiences as valuable…  
The availability of meaningful options depends on access to a societal culture, and on understanding the history and language of that culture… (p.84) 
Kymlicka elaborates on what he means in Politics in the Vernacular: 
My basis argument can be summarized this way: modern states invariably develop and consolidate what I call a societal culture – that is, a set of institutions covering both public and private life, with common language, which has historically developed over time on a given terrirory, which provides people with a wide range of choices about how lead their lives…   
These societal cultures are profoundly important to liberalism, I argue, because liberal values of freedom and equality must be defined and understood in relation to such societal cultures. Liberalism rests of the value of individual autonomy – that is, the important of allow individuals to make free and informed decisions about how to lead their lives – but what enables this sort of autonomy is the fact that our societal culture makes various options available to us. Freedom, in the first instance, is the ability to explore and revise the ways of life which are made available by our societal culture. (p.53)
In The Limits of Nationalism, Gans says: 

People have a fundamental interest in adhering to components of their identity. They have an interest in being respected for their identity and the components that comprise it, or at least in not suffering humiliation or alienation because of it…  desires involving objects in which people have fundamental interests must be given special weight in determining the contours of the political order. …such desires are very different from desires involving objects in which people do not have fundamental interests (for example, the wish to spend a vacation on one particular island as opposed to another island)… [With regard to] people ’s interest in adhering to components of their identity, the assumption that it is fundamental has ample support. It is supported by the fact that many people regard themselves as having this interest, especially in relation to their cultural identities (p.43). 

Terrorism

Terror and Consent - Philip Bobbitt / The Missing Martyrs: Why There Are So Few Muslim Terrorists - Charles Kurzman 

Perhaps no book has had as much of an impact on my views than Terror and Consent. It provides the building blocks for my views about the importance of Western power, the need for carefully managed interventions and the importance of international alliances with autocratic and multilateral institutions. Bobbitt’s book has insights into:

  1. Why the argument that “terrorism isn’t as common as dying from peanuts / fridges / insert your analogy here” are wrong. 
  2. Why the nature of terrorism has changed: we’re no longer looking at nation state terrorism, but market state terrorism. 
  3. Why the nature of war changing has implications for law enforcement and military strategy (i.e., why domestic police action is not sufficient for stopping terrorism). 

To elaborate on that final part, the failure in Iraq between 2003 and 2006 can be attributed to a failure to understand the changing nature of war. Bobbitt correctly attributes the chaos in the aftermath of the war to the inadequate number of troops required; and the de-Baathification of the Iraqi army. From a “law enforcement” only view, it made sense not to send more troops after the regime had been dismantled. But market state terrorists are not conventional armies, they thrive off security vacuums. As Bobbitt explains: 

[Only 10,000 troops were in a position to patrol Baghdad.] By comparison, New York City has a force of 38,000 police and it, despite some depiction in the movies, is not a war zone. But then neither was Baghdad, in a conventional sense, the war was over. The looting and civil disorder that followed were [inappropriate not considered]  tasks for the military…  

Coalition troops went from hot spot to hot spot , each of which they had to abandon to get to another… It needn’t have been this way, Based on recent experience in the Balkans, a number of studies had made remarkably similar estimates as to the number of troops that would be required to subdue Iraq. NATO had deployed 200 soldiers for every 10,000 civilians. For Iraq, this rough ratio would have meant a stabilisation force of more than 450,000 - three times the number commanded by the Coalition (p.164-165). 

The issue in Iraq was not that there wasn't a plan, but that the plan was based on a misconceived idea of war in the 21st Century. The ideas about having more agile, flexible militaries which rely on modern technology such drones is part of this problem: Rumsfeld’s view about how many troops were needed did not come out of nowhere, they were based on believing in the predominance of air power. But this view poses a problem because of the security vacuums that get created. 



By way of analogy, the criticism made about information sharing between the CIA and the FBI in the 9/11 Commission Report were based on the idea that we could make a clear distinction between law and strategy, between law and war. You cannot: we are at war with a new force that requires us to blur the lines: wars require maintaining civil order and crimes require military responses. 
 
I have written a review of the Missing Martyrs so wont say much more about it here


Intended Consequences

Engineering the Financial Crisis - Krausz and J Friedman

I think Jeffery Friedman is one of the most interesting living thinkers. He makes me think really hard about almost everything he writes about. His book on the financial crash puts the cause of the crash to minimum capital regulatory requirements. I think his explanation of the crash is persuasive  but its even more important for how I view conflicting viewpoints and public policy. 

Briefly, he shows that the capital adequacy requirements led banks to indulge in mortgage backed securities, because it helped them meet their capital requirements more quickly. This was not an act of irrationality, it was not an act of malicious greed: bankers themselves invested in these personally so it cant be said that they expected them to fail. But that’s not the main reason why I love the book and how its been influential. Here are some relevant extracts: 

Specifically, the vulgar usage of ‘‘irrationality’’ treats people’s errors as inexplicable... Another way to put the point is that the vulgar usage does not recognize that reality is complex and can produce many different interpretations, each of them plausible to sane, rational people.... Thus, the vulgar notion of irrationality treats accurate and representative ‘‘information’’ about objective reality as if it were there to be had, sans interpretation, by any rational agent, such that behavior that does not take our retrospective interpretation of this information into account must have been due to a lack of rationality—rather than to an absence of accurate, representative information, or the presence of an incorrect interpretation of such information…. To make an error in logic, let alone to be unaware of a fact, is not the same thing as to be irrational (p.50-51) 
…we believe that Akerlof and Shiller’s critique of economism is founded on an error in logic: a conflation of irrationality with error. One can be perfectly rational and as unemotional as Star Trek’s Mr. Spock, yet make a mistake in thinking that a given action will, indeed, be instrumental to one’s desired end (p.125). 
This has incredibly important implications. First, I treat people who make conflicting political arguments as though they are rational. I may believe they are wrong, but I try make sure that its extremely rare where I believe someone is being unreasonable. My view on the minimum wage for example, is that its bad for employment, bad for prices, and reduces immigration which is bad for productivity and GDP. I don't hold, for example, Owen Jones as being unreasonable because he has a different view. The idea is linked to the premise that the political should not be personal or emotional. It should be a dialogue about competing ideas. 

Second, and Friedman probably wouldn’t agree, it has implication for democratic government. The Caplan-esque argument is that the typical voter is “irrational” and so shouldn't be trusted. But Caplan here is making the same mistake: wrongness is not the same as rationality. The complexity that we see should lead us to being humble about certain views. Again, we are not talking about how right an idea is, we are talking about how to deal with it. 

Third, the propensity to attribute malicious motives to people with different views (“he’s funded by Exxon” or “he’s evil”) is a weakness in political thought. It assumes you've already made the jump from X-view to malice without appreciating that the person with a different view could just be wrong. It assumes there is no reasonable reason for disagreeing with you. I voted Remain, but I wouldn't attribute evil, malicious funds or even irrationality to Leave voters without more evidence.  

Methodology

Language, Truth and Logic - A J Ayer / Reconstruction of Automobile Destruction: An Example of the Interaction between Language and Memory, Elizabeth Loftus and John Palmer, Journal of Verbal Learning and Verbal Behaviour (1974). 

I have spent a number of posts explaining why I have a strong aversion to narrative journalism. Journalists will write entire articles without mentioning a single study - and it really irks me. In my posts, I try to make sure that every single statement is backed by a study, and I draw people’s attention to anything that is not supported by data. This view comes from my early adherence to logical positivism. 

I’m a bit of a fundamentalist when it comes to this - I have said that if there isn't a study in your article, it probably shouldn't be published; I even held the view that personal testimony should be removed from court rooms. The unreliability of eye witness testimony, the claims of the religious to have had personal encounters with ghosts, Gods and other assorted gremlins belies its importance. Loftus’ study was the first study I read during my A Levels that led me to this view; subsequent meta-analysis have confirmed their findings. 

I recently had a discussion with Sam where he disagreed with this view - his view was that where there was no data, anecdotes could be used. I disagree. Conflicting anecdotes, the infallibility of the human mind, the fact that n=1, and biases lead me to refuse to give any weight to any anecdotes.

Other Economic
  1. The Great Recession: Market Failure or Policy Failure? - Robert Hetzel
  2. ‘Mark Sadowski on fiscal austerity, with and without monetary offset’ - Scott Sumner, Money Illusion
  3. The Use of Knowledge in Society - Hayek
  4. Are Banking Crises Free Market Phenomena? - George Selgin
  5. 7 Reasons Not to Care about High Pay - Sam Bowman, ASI Blog
  6. Mammon and the Pursuit of Empire: The Economics of British Imperialism, Davis and Huttenback
  7. Redefining the Poverty Debate, Kristian Niemietz
 
On Inequality
  1. Overstating the Costs of Inquality, Scott Winship, National Affairs
  2.  It's the Market: The Broad-Based Rise in the Return to Top Talent, Kauh and Kaplan, Journal of Economic Perspectives
  3. Ben’s stuff: Why Does the Son Rise? - ASI Blog; Why do rich parents give birth to rich kids?; ASI Blog.
  4. Q&A: A Sociologist on Inequality, New York Times; and Income Equality: A Search for Consequences, New York Times
  5. Capital - Piketty
  6. 'Why We're in a New Gilded Age' - Paul Krugman, New York Review of Books; Review of Capital, Mervyn King, The Telegraph and 'Capital Punishment', Tyler Cowen, Foreign Affairs.
  7. Capital Taxation in the 21st Century, Auerbach and Hasset, NBER Working Paper
  8. Does housing capital contribute to inequality? A comment on Thomas Piketty’s Capital in the 21st Century, Bonnet et al, Science Po Economic Discussion Papers
  9. Is Piketty’s “Second Law of Capitalism” Fundamental?, Krussel and Smith, Journal of Political Economy
  10. A note on Piketty and diminishing returns to capital - Matthew Rognlie
Other 
  1.  Wait a minute: is the government self-interested or isn't it? - ASI Blog
  2. Taxes, Lawyers and the Decline of Witch Trials in France, Noel Johnson and Mark Koyama, Journal of Law and Economics, Vol. 57 
  3. Wishful Thinking or Buying Time? The Logic of Appeasement in the 1930s - Norrin Ripsman and Jack Levy 
  4. The Threatening Storm - Kenneth Pollack
  5. 'Saddam, Israel, Nuclear Alarmish Justified?', Brands and Pallki, International Security, Vol. 36
  6. 'Krugman's Response to Alex', Tyler Cowen, Marginal Revolution
  7. The Case for Israel - Alan Dershowitz and the masterful takedown Beyond Chutzpah - Norman Finkelstein; One State, Two State - Benny Morris; Righteous Victims - Benny Morris; most scathing reviews written by Benny Morris.

Thursday 28 July 2016

The Empirics of the Places We Go Part I: Economic Effects of Immigration


Congratulations!
Today is your day.
You're off to Great Places!
You're off and away!
You have brains in your head.
You have feet in your shoes.
You can steer yourself any direction you choose.
You're on your own. And you know what you know.
And you are the guy who'll decide where to go.
 
Oh the Places You’ll Go by Dr Seuss

Believe it or not, this post was supposed to be about terrorism. But as I began writing, I realised that I couldn’t make any policy prescriptions on the basis of the link between immigration and terrorism without considering the wider effects. This is a post about immigration and why, at least for now, we should keep yelling for more immigration. I cannot provide a figure and I don’t know when we’ll have to stop (and we may well have to stop one day), but we are nowhere near.  
I have become less pro-immigration as I have written this post. I don’t want to give the wrong idea, compared to the average person, I’m probably extremely pro-immigration but I’ve definitely become less certain and more restrictive than I thought I was. I’ve read quite a lot for this post but if you feel I’ve missed anything, please let me know.  Here is a structure of the essay below:
  1. Shunning Sylvester McMonkey McBean: Economic Effects of Immigration  
  2. The Predicaments of a (Douglas) North-Going Zax: Institutional Effects of Immigration
  3. The Honourable Fifth Column: Crime, Terrorism and Immigration
  4. Defining the Priors of Star Bellied Sneeches: The (Tenuous) Compatibility of Liberal Nationalism and Utilitarianism
  5. How To Tell a Klotz from a Glotz: A Conclusion on Public Policy and Uncertainty
This series will likely be in two or three parts, I’m not sure how the above will be divided exactly yet. There are end notes which aren’t integral to the overall argument at the bottom of the post so feel free to ignore those (in this part they cover irrelevant things like a non-nationalist justification for the existence of State of Israel, the signalling model of education, globalisation’s effect on workers in high-wage countries and voting Leave, and the Chinese hukou system). Section 1 does not cover institutional arguments that could affect economic output, this will be handled in section 2.
Thanks to Sam for giving me some thoughtful comments on an early draft of this post. Any mistakes are his, and his alone.
1. Shunning Sylvester McMonkey McBean: Economic Effects of Immigration  
It is a truth universally acknowledged that the British public must be in want of less immigration, a fundamental value of 21st century Britain:  
There are four reasons why this truth should not be overstated – and, in my view, should not be heeded. First, it’s important to decompose the results. According to a recent Ipsos Mori poll, when you ask whether immigration has affected them personally, a slim majority (51%) say no effect and further quarter say it’s affected them positively. As Chris Dillow notes, ‘whilst most people think immigration is a national problem, they don't believe it to be one in their own area.’
\\bdb-law.co.uk\usrdata\MyDocs\MLA\Downloads\BDOfbSFCYAAHqhk.png
Moreover, the areas with the highest levels of immigration, are the most pro-immigration (see graph above from Markaki (2012)). The less experience you have of immigration, the more you are likely to oppose it. This is not caused by white flight (i.e., white people who don’t like immigrants moving away from high immigration areas)[1]  

Second is that the reasons people are anti-immigration matter – especially if, like me, you buy into the Burkean trustee model of representation. The table below shows the results from a recent ComRes poll:
https://ton.twitter.com/1.1/ton/data/dm/755138493618282499/755138493647552513/XLDb3GDQ.jpg
This interesting not least because it is strong evidence that anti-immigration views are based on the economy, rather than perceived socio-cultural changes or racist concerns.[2] 68% of people are not willing to reduce their income at all even if meant reduced immigration. It also means that, just like swathes of people in favour of nationalisation, if you buy into the trustee model of representation, MPs can legitimately ignore their views if immigration is actually good for the economy.
This is linked to the point that people may be anti-immigration because theyre wrong about the economic effects. Indeed, when people are made aware of the benefits, support for immigration rises as noted by The Guardian:
A YouGov poll found that the proportion of those questioned who viewed immigration negatively dropped from 63% to 54% after they were told that the government's financial watchdog believes that higher immigration will help the economy grow and ease pressure to cut spending…
YouGov found that, after being told the "full facts", the proportion of respondents holding positive views about net migration grew from 32% to 39%. The numbers wanting an end to all immigration dropped from 16% to 12%, and those advocating a small number of skilled migrants fell from 47% to 42%.
Third, we shouldn’t underestimate the value of the trade-off. The public may not like immigration, but they may vote for it because of the other things it brings. Brexit has given us an opportunity to see an example of this.[3] Despite high levels of opposition to more immigration, people are willing to forego this in exchange for access to the single market (source, see also here and here):
https://pbs.twimg.com/media/Cm_PsnQXgAAzGle.jpg
Fourth, relates to the perception of numbers. Sam noted a study which showed that ‘most people know very little about the extent and direction of income inequality in their societies, or where they fit in to the income distribution.’ Sam went onto argue that, as a result, ‘we can only solve [this problem] by making people feel less unequal – not by making them less unequal in fact’ because perceptions are unhinged from reality. I think it’s a good argument and it can be applied to immigration.
We can see from the graph above (source) above that in the UK, the perception is that just over 30% of the population are immigrants. The reality is that around 12% are. The perceptions are far from reality – given that immigration curbs may be costly and we cannot rely on people’s perceptions about the levels of immigration, we need not give premium to people’s preferences – especially given the aforementioned three points.[4] A more PC argument that Sam might make is that we need to inform people more about immigration levels – either way, the lust for lower immigration doesn’t carry much weight with me.
1.1 Domestic Economic Impact
I’m not convinced that there is a good economic case against high levels of immigration. For an overview of why, Lisenkova et al (2013) look at what would happened if we really did reduce migration to the “tens of thousands.” Their model splits people up into foreign born and native born. These two groups typically have different wages, consumption and life spans (that impact on the former two). They find significant negative effects for natives:
http://www.niesr.ac.uk/sites/default/files/images/Blog/katerinablog.png
...by 2060 in the low migration scenario aggregate GDP decreases by 11% and GDP per person by 2.7% compared to the baseline scenario... this policy has a significant negative impact on public finances, owing to the shift in the demographic structure after the shock. The total level of government spending expressed as a share of GDP increases by 1.4 percentage points by 2060. This effect requires an increase in the effective labour income tax rate for the government to balance its budget in every period. By 2060 the required increase is 2.2 percentage points
The results are conservative in sense that they model for just below 100,000 (I’m not sure if this is what the Conservative Party means when they say they want immigration in the “tens of thousands”) and they don’t take into account the effects of immigration on things like total factor productivity. These results are fairly representative of the literature but its worth considering in details the specific impacts of immigration on employment, wages and public finances.  
On employment, Luchino et al (2012) find ‘no association between migrant inflows and claimant unemployment.’ The correlation actually goes the other way so that a 2 percentage point increase in the amount of migrants leads to a 0.02 percentage point reduction in the claimant (i.e., unemployment) count but this result is not significant. They go one step further and look at whether this association changes depending on the business cycle – still no association.  Note that this study was looking at the UK – but when you look at cross-sectional evidence, you find the same thing (see Figure 2 below).
https://img.washingtonpost.com/wp-apps/imrs.php?src=https://img.washingtonpost.com/blogs/wonkblog/files/2014/05/Screen-Shot-2014-05-06-at-7.20.08-AM.png&w=1484
The literature does raise some issues about whether this positive effect extends to everyone. For example, Dustmann et al (2005) find no significant effect on the overall level of unemployment but they did find it had a negative impact on those with only O-levels but a positive impact for those with A levels or university degrees. I’m not persuaded by this line of research for two reasons. First, when you look at the impact over long periods of time, the effect diminishes.  As Devlin et al (2014) note:
... our assessment is that there is relatively little evidence that migration has caused statistically significant displacement of UK natives from the labour market in periods when the economy has been strong... The evidence also suggests that where there has been a displacement effect from a particular cohort of migrants, this dissipates over time – that is, any displacement impacts from one set of new arrivals gradually decline as the labour market adjusts
And this ties quite nicely with a similar but quite distinct point: the impact of immigration tends to be tamed by how flexible your labour markets are. As Sam says ‘the more rigid a labour market, the worse immigration is for native workers.’ Looking at European data, Angrist and Kugler (2003) show that ‘increasing the severity of labour standards by one standard deviation, about the difference between Denmark and Belgium, would increase the negative effect of immigration from -0.027 at the median to about -0.042.’  Just to really hammer the point, the OECD produces this helpful graph:
C:\Users\Mustafa\Desktop\download.png
The countries on the right hand side are likely to be integrating migrants into their labour markets better than those on the left. I decided to look at how these countries tacked on the Heritage Foundation’s measures of Labour Freedom (100 being the highest levels of labour freedom and 0 being the lowest). The average score for the ten countries with the highest risk is 56. The average score for the ten countries with the lowest risk is 70. That is a significant difference. You can check my working out by downloading the data here.
Fortunately in the Anglosphere, where the average Labour Freedom score is 81, we do have dynamic labour markets which moderate the effect – and, if anything this should lead you to favour economic freedom rather than restrictions on immigration. Why would flexible labour markets matter? The obvious is that people who do lose jobs can more easily find jobs in a labour market with less minimum wages, hiring and firing regulation and centralised collective wage bargaining etc. Di Tella and MacCullock (2005) find that if France had a labour market as flexible as the U.S., ‘its employment rate would increase 1.6 percentage points, or 14% of the employment gap between the two countries.’
But there’s an even more interesting mechanism that I wasn’t aware of until I read Brian and Kovak (2016): immigrants respond to labour market demands in a way that equilibrates the labour market. How? Well, natives don’t appear to be as geographically responsive to labour market demands: if employment declines and you lose your job, as a native, you are (relatively) not likely to get up and go somewhere where there is demand. But that is exactly what immigrants do: they will get up and go (see Figure 1 above). And this makes a huge difference for natives:
We find that in metro areas where the Mexican-born comprised a substantial share of the low-skilled workforce prior to the Recession, there was a much weaker relationship between labour demand shocks and native employment probabilities than in areas with relatively few Mexican workers. Natives living in metro areas with many similarly skilled Mexicans were thus insulated from local shocks, as the departure (arrival) of Mexican workers absorbed part of the relative demand decline (increase).
Changes in employment probabilities for natives living in cities with large Mexican populations are much less related to local demand conditions than are changes in cities with few Mexicans. The relationship in above-median cities is 61 percent weaker than in below-median cities. Thus, native employment probabilities were insulated from local shocks in the presence of substantial numbers of Mexican-born workers, with improved native outcomes in the hardest hit cities and diminished ones in more favorable markets
The key section to look at in Table 5 above is Panel B. You can legitimately question whether this result should be applied outside of the U.S. or Mexican immigrants, but it’s a fascinating study which shows the economic benefits of immigration. This effect seems to apply to “refugees” as well. I will be discussing the literature on refugees and native wages a little further below, but just in terms of outcomes on employment levels, Foged and Peri (2015) is one study. They conclude:
... the increase in refugee-country immigrants pushed less educated native workers to change occupation. This move was significant and towards non-manual occupations, and particularly strong when workers changed establishment.
The idea is that they were ousted by immigrants but then pushed into better jobs. Indeed, ‘an increase in refugee-country immigrants by 1 percentage point of employment increases the complexity of native jobs between 1.3 and 3.1%.’ Just to clarify, in this section I’m talking about the overall, net employment effect (even for those at the bottom). There are some who will lose out. One recent study undertaken by the Federal Reserve of Boston finds that for every foreign nurse hired in city in California, one to two fewer native nurses are employed in that same city (Cortes and Pan (2014)). As should be clear, the general effect on employment to me seems to be positive.
On wages, much of the general conclusions above relating to employment apply: generally small positive effects (Ottaviano and Peri (2012), Longhi et al (2005)) and Wadsworth et al (2016)) – the graph below is taken from the latter).
There are some studies which find negative average effects but these don’t seem to withstand scrutiny. The most famous is Borjas (2003) which found a 3.8% decline in wages. When you include an expanded data series, this drops to 2.2%. Even more interestingly, Borjas uses a weekly wage variable. Brauw and Russell (2014) multiple the weekly wage by the amount of weeks actually worked and then re-run the regression. You’d think it shouldn’t make a difference – but it does: its no longer statistically significant! The finding that there are modestly positive effects on average wages, therefore, seems to stand.  
However, this average increase in wages potentially impacts people in different parts of the distribution in different ways. I’m not sure I can do any better than Flipchart Rick’s post “It would be surprising if people didn’t worry about immigration.” I recommend that everyone read this post, and everything ever written by Rick. The Migration Observatory notes that:
The effects of immigration on workers within specific wage ranges or in specific occupations are more significant. The greatest wage effects are found for low-waged workers. Dustmann et al (2013) find that each 1% increase in the share of migrants in the UK-born working age population leads to a 0.6% decline in the wages of the 5% lowest paid workers and to an increase in the wages of higher paid workers. Similarly, another study focusing on wage effects at the occupational level during 1992 and 2006, found that, in the unskilled and semi-skilled service sector, a 1% rise in the share of migrants reduced average wages in that occupation by 0.5% (Nickell and Salaheen 2008).
There are some interesting studies on the effect of refugees on wages. The main studies look at Cuban refugees in Miami after the Mariel Boatlift in 1980 (over 120,000 Cubans arrived in Miami). Here is Card’s (1990) headline results:
Apart from the temporary increase in relative wages of workers in the lowest quartile between 1979 and 1981,the distribution of non-Cubans' wages in the Miami labor market was remarkably stable between 1979 and 1985 .. the influx of Mariel immigrants had virtually no effect on the wage rates of less-skilled non-Cuban workers. Similarly, there is no evidence of an increase in unemployment among less-skilled blacks or other non-Cuban workers. Rather, the data analysis suggests a remarkably rapid absorption of the Mariel immigrants into the Miami labor force, with negligible effects on other groups.
Enter Borjas (2015). He says that the Card study didn’t measure the right group that was likely to be affected by the influx of Cuban refugees (i.e., those without a high school education). When looking at this group of people, he found that wages were depressed 10-30%. I’m inclined to believe that Card’s finding is more robust for the following reasons:
  1. Borjas’ study used four cities as controls: if every state (including those that did not experience an inflow of Ciban refugees) did not have a comparable fall, then it makes sense to attribute the decline to the immigrants. Which brings us to Peri and Yasenov (2015) - they use the ‘synthetic control method’. To get the point: the method allowed them use a larger pool of control states (43, rather than 4) – and when using this larger pool, they find that ‘the change in wages and unemployment of Miami highschool dropouts relative to the control group in 1979-1982 was, by no means, unusual andwell within the distribution of other cities’ idiosyncratic variation.’
  1. Borjas’ sample consists of 17 to 24 workers in Miami each year. That is ridiculously small, and makes me question whether the results can really be statistically significant: his findings rely on wage falls for 16 people. Card’s sample covers people aged 16-61, Borjas’ covers 25-59.  Peri and Yasenov note that the smaller CPS dataset (used by Borjas) rather than the larger ORG dataset (used by Card) the measurement error for average log has a standard deviation of 0.15 log points – and so 10-30% depression of wages in the data really is insignificant.
  1. When using a Synthetic Miami vs Miami model using the ORG dataset, Peri and Yasenov find that there was ‘no significant difference in the post-1979 labor market outcomes between Miami and Synthetic Miami (the control group). Neither wages (annual, weekly or hourly) nor unemployment of high school dropouts differ between Miami and the control after 1979, up to 1983.’

lwageTCMeanORG1
 
  1. Card includes women, Borjas excludes them. When David Roodman applied Borjas’ model to women, he found the opposite result (see graph above showing the rise in wages). And when you combine the data for men and women, you find that its essentially a flat line (see graph below).

lwageTCMeanORG3
  1. As Sam points out, ‘the relative wages of high school dropouts recover entirely by 1990 – the effect Borjas has found only holds in the short-run’ (more on this below).
  1. The aforementioned Brauw and Russell make another significant point: large scale immigration is a mass entrance of new people into the labour and accordingly, essentially, a labour shock. They look at the largest labour shock: the entrance of women into the labour market in the U.S. post 1960s.  And their results don’t match Borjas’ (2003 or 2015) model. They find that ‘insignificant coefficients on weekly wages among men and positive, statistically significant coefficients on annual wages.’ It is more likely that Borjas’ findings are wrong than there being a discrepancy in labour market shocks
  1. The preponderance of the limited evidence that we have on refugees seems to confirm Card, rather than Borjas. Foged and Peri (2015) (already mentioned above), for example, looked at refugees arriving in Denmark between 1986 and 1998. We have seen that low-wage earners were pushed into higher paying jobs. This explains their other significant finding: there is a 1 to 1.8% native wage increase following a 1 percentage point increase in share of immigrants from the refugee-sending countries studied.  
Borjas has responded to some of these claims in an article for the National Review. His response is largely unpersuasive. What he calls the most egregious flaw in Peri and Yasenov is including 16-61 year old because ‘your 16-, 17-, or 18-year-old son or daughter who is now a sophomore, junior, or senior in high school is classified as a high-school dropout because he or she does not yet have that diploma.’ That seemed to me to be a fair criticism – except it only accounts for about 20% of the increase in Peri and Yasenov’s study. I e-mailed Peri about this and he confirms what I suspected:
The key point is that none of his point really matters for our results of no effect on wages. We now [in a working paper due to be published soon] drop all workers below 19 in our main sample and the results are totally unchanged… when one uses the larger, more reliable sample, with lower measurement  error which is the  May-ORG sample, one never finds the impact on Miami less educated wages.
Only with SOME samples in the much smaller March-CPs data, which rely on 15-20 observations per city in the relevant group (and so massive measurement error) one can find some negative effects on some subgroups. This is why Borjas is so obsessed in the sample choice. Only in one small sample, only with March-CPS data one can find a negative impact. But given the large standard  error of those averages that occurrence can be by pure chance.
The aforementioned Wadsworth et al study of EU immigration finds ‘no apparent link between changes in the real wages of UK nationals and changes in EU immigration’ – even for the least skilled native workers. Figure 11 below shows the effect on real wages as against the change in immigration levels from the EU. So much for the negative effect on wages of the lower rung of the economic ladder.
http://static1.squarespace.com/static/56eddde762cd9413e151ac92/t/57349d581d07c02082ddba72/1463065955716/?format=1000w
Even if you wanted to maintain that there is a negative effect for low-earners, the studies that find negative results really are minimal. Nickell and Saleheen (2015) looking at the impact of a 10 percentage point rise in the proportion of immigrants working in semi/unskilled services, find a 1.88 percent reduction in pay (see Table 7). What does this actually mean? Here’s how Jonathan Portes explains it:
Well, the first thing to note is that a 10 percentage point rise in the proportion of migrants working in a sector – the amount needed to generate the “nearly 2 percent” wage impact is very large.  Indeed, it is larger than the entire rise observed since the 2004-06 period in the semi/unskilled services sector, which is about 7 percentage points…
Moreover the estimated impact is partly simply a compositional one – reflecting the fact that migrants earn less, as well as the impact on native wages. Allowing for this, we can calculate that the new paper implies that the impact of migration on the wages of the UK-born in this sector since 2004 has been about 1 percent, over a period of 8 years. With average wages in this sector of about £8 an hour, that amounts to a reduction in annual pay rises of about a penny an hour.
But here’s another interesting reason why you shouldn’t be that concerned about the wage effects. Those low-earners whose wages are probably affected are themselves immigrants. Manacorda et al (2012) find that ‘immigration has primarily reduced the wages of immigrants - and in particular of university educated immigrants - with little discernible effect on the wages of the native-born.’ Studies that look at wage effects without disaggregating groups are borderline worthless: Ottaviano and Peri (2012) find that there are positive wage effects for average native works (including those without a high school degree) in the range of 0.6 to 1.7% - but there is a 6.7% decrease for existing immigrants.
Even more important is that the effect on wages seems to be short lived (see above, and Devlin et al). As far as the compositional effects on the poorer people go, I have to say quite bluntly: even if there is an effect (unlikely) and it’s not short term (unlikely):  you can be worried, but don’t expect the rest of us to heed your concerns[5]. There are more significant reasons why I believe this that will be handled below, but even just looking at the domestic side of things, I can’t put it any better than Portes:
Yes, we're doing this for the good of our country, and yes you may lose out, but ultimately we still have to do this… Just as we said to the coal miners 30 years ago: 'Sorry we can get our coal a lot cheaper abroad. We can't afford to keep on propping you up.'… In the case of coal mining and immigration relatively liberal policies had benefited people, on average, even if some individuals had lost out, and that these policies had therefore been broadly sensible: the human consequences, overall, were positive.
On public finances, the Office of Budget Responsibility (2013) looked at the question of how much immigration affects debt levels and finds what the Lisenkova et al (2013) found:
By 2062, our debt as a percentage of GDP will be 175% with no immigration, it would be 100% with 140,000 a year and 75% with 260,000. There are two reasons for this. First, immigrants just don’t drain resources: they pay taxes, they create wealth. It is little surprise that the OBR has said that for Chancellor Osborne to reach his deficit target he must miss his target to reduce net migration. Indeed, revising immigration up from 105,000 to 185,000, led to the OBR revising output up by 0.9% (see this Guardian report). As the Financial Times noted in 2014, migrants ‘created one in every seven UK companies’ and ‘are responsible for creating 14 per cent of British jobs.’  The report goes onto say:
... entrepreneurial activity among the migrant community was nearly double that of UK-born individuals: 17.2 per cent had launched their own businesses, compared with 10.4 per cent of those born here. They are also, on average, eight years younger than indigenous entrepreneurs at 44.3 years old, compared to 52.1. This is despite the extra challenges they face, including access to finance and cultural and language barriers
This finding is not unique to the U.K. As Anderson (2016) notes, in the U.S., immigrants have ‘started more than half (44 of 87) of America’s startup companies valued at $1 billion dollars or more and are key members of management or product development teams in over 70%’ of them. The value of just these 44 companies is ‘$168 billion, which is close to half the value of the stock markets of Russia or Mexico.’
The second reason why the effect on public finances shouldn’t be surprising is because the pull of welfare states is often overstated. Evidence from the U.S. seems to suggest that the welfare state is really not a magnet or, at the very least, immigrants are not using welfare states any more than the native population.
…immigrants generally participate in the safety net at lower rates than natives once we restrict ourselves to comparisons within the set of lower‐income families. This is true for almost all programs we consider and is true both before and after [the 1996] welfare reform [which restricted welfare for immigrants]… our results show that immigrants rely more on earnings as a source of income (than do natives) and the degree of reliance has increased post‐welfare reform
On the question of immigrants coming to Western countries to use our welfare states and draining them, the general argument, therefore, seems to fail. However, there may be a differential depending on the origin of the immigrant. Which brings us to my favourite Radio 4 programme, More or Less. In an episode from 2014, they investigated the claim that immigrants create public debt. They note that:
…between 1995 and 2011, on average each European immigrant put about £6,000 more into the public purse than they took out. Non-European immigrants, on the other hand, each took out about £21,000 more than they put in. on average, each native Briton took out roughly £11,000 more than they put in between 1995 and 2011.
FullFact (an under-appreciated site) has a nice summary of the research on the fiscal effects of immigration. I suggest you ignore the studies put out by MigrationWatch because their methodology is so flawed, I barely know where to begin. But, even then, the broader picture is precisely this: European immigration seems to be okay for public finances, non-European immigration does not.
Table 2
It is worth considering the Dustmann and Fratitini research in a bit more detail. As you can see, they find that non-EEA migrants cost the public purse. However, when you take into account that some public expenditures are fixed, even non-EEA migrants outperform natives when it comes to contributions to the public purse:
Aside from the origin of the immigrant, their reason for coming may matter: refugees, for example, may be less likely to be able to work because of their age and/or skills and therefore their contributions to GDP and to public finances may be smaller. The evidence here does seem to show a negative impact on public finances. Ruist (2015) looks at the effect of the recent refugee flow into Sweden on public finances. The table below is worth considering:
http://www.voxeu.org/sites/default/files/image/FromMay2014/ruist%20table1%2027%20jan.png
Ruist finds that refugees account for 5.6% of public spending. Refugees in Sweden constitute 5.1% of the population so this isn’t hugely disproportionate. But, when you look at the table, you see that they account for 55.4% of social assistance and 25.3% of the spending on “crime and justice.” Refugees are a drag on public finances: the net fiscal redistribution is 1% of Sweden’s GDP. Ruist explains this finding:
... they performed much worse than the rest of the population on the revenue side, where they contributed only an estimated 3.4% of total public revenue, essentially through direct and payroll taxes. The reason is clear: the employment rate among adult refugees was 20 percentage points lower than that among all adults. The reasons include poor language skills, lack of applicable training, lower female labour force participation rates, and so on... Four-fifths of the redistribution was due to lower public per capita revenues from refugees compared with the total population, and one-fifth to higher per capita public costs.
There still may be a drag but Ruist’s finding that around a pretty insignificant 1.35% of GDP being redistributed should dampen the weight of this argument. The more important point about Ruist’s finding relates to the labour market. With flexible labour markets, refugees can be integrated into employment far quicker – this has been addressed above, but when you put the points together you get a simple maxim: If you want immigrants not to be a drag, free your labour markets!
If you only cared about the public finances side of things, you should have no qualms, at the least, with European (and, let’s be honest, Anglosphere) immigration but you might want restrictions on non-Europeans. Leaving aside the impacts on public finances in the past, the OBR’s model is clear: you should want to encourage immigration to a level near 260,000. You might even cut out the nationality element and prefer that immigrants must have (1) a job or (2) a high paying job. This policy prescription has to be weighed against further factors below.  
The public finances argument is not that conclusive for me (GDP, GDP per capita and wages matter more in my opinion) but here’s an interesting idea. If you do care about public finances and the impact from non-European immigrants – why not say you are anti-legal immigration but you are pro-illegal immigration? Why not support worker programmes? I do not support these and I will explain why and return to the issue of public finances below. But, as we’ll see, the benefits to immigrants are so large, that the answer should almost never be restriction of immigration but changes of other policies.
As a concluding point to this subsection, the above broadly positive economic findings are consistent and significant. Ed West’s book The Diversity Illusion is an anti-immigration book worth reading (full disclosure: I consider Ed a friend) but it does contain a common incorrect refrain of the anti-immigration writers when he says ‘certainly some economists believe that immigration benefits the economy… but just as many economists argue to the contrary’ (my emphasis) (p.73).
This is not true. The weight of the evidence is behind the positive effects of immigration (a point that Ed accepts when he says that ‘economist tend to approve’ of immigration on p.84). There is a better argument that these effects are closer to zero and only marginally positive, but that doesn’t seem to be the argument of most anti-immigration writers. Below is the IGM Economics Experts Panel view of immigration’s effects on the average worker.
10% is not “just as many” as 63%. As accepted, there will be losers but we cannot make policy on the basis that there are some losers. I use the word “some” very specifically. Aubrey et al (2016) look at the labour market effects (employment, wages etc.), the fiscal effects and what they call ‘market size effects.’ Market size effects are aimed at measuring ‘aggregate demand for goods and services in the receiving and sending countries’ as this ‘determines firms’ entry and exit decisions and in turn, the numbers of entrepreneurs and goods available to consumers’. They then try to measure who exactly benefits on the basis of these effects:
The set of winners represents 6 9.1% of OECD non-migrant population aged 25 and over. This share increases to 83.0% if one considers the 22 countries whose GDP per capita was above USD 30,000 in the year 2010. Contrary to popular perceptions, winners mainly reside in net immigration countries; their gains can be important and are essentially due to the entry of immigrants from non-OECD countries, which has a drastic effect on market size… The average effect on non-migrants is positive in 28 OECD countries, nil in France, and negative in 5 traditional countries of emigration.
In the second figure above (labelled (b)), the average effect of migration from outside the OECD on natives within the OECD are considered. You can plainly see that ‘effect of extra-OECD migration is positive in 22 countries.’ How does this compare with intra-OECD migration? Aubrey et al identify ‘17 winners and 16 losers (the effect is nil in Sweden)’. Their main conclusion on total welfare is as follows:
Overall, extra-OECD migration flows increase the average utility of non-migrants by 1.2% in the OECD (and by 0.9% if older cohorts of migrants are included in the average)... Intra-OECD migration flows decrease the average utility of non-migrants by 0.1% in the OECD. Hence, the bulk of welfare gains from global migration is driven by extra-OECD migration, in line with Di Giovanni etal. (2015) or Iranzo and Peri (2009).
This is not me picking and choosing studies, the bulk of the economic literature and economists support the idea that immigration is a good for the economy and the domestic state with minimal, short-term downsides. Anti-immigration activists would do better to make a cultural or institutional case – something I’ll consider in the next part.
1.2 International Economic Impact
Of course, the domestic impact is not all we care about –  the impact on global economic outcomes is similarly important. The different weight we should give to domestic and international factors will not be explored in this part. The first impact to consider is the net impact on GDP. The most famous paper is Clemens (2011) in which he states:
For the elimination of trade policy barriers and capital flow barriers, the estimated gains amount to less than a few percent of world GDP. For labor mobility barriers, the estimated gains are often in the range of 50–150percent of world GDP
Clemens notes that if you assume that the average gain for an immigrant is $7,500pa (which is highly conservative), then you would have overall gains of 20 to 60% of global GDP. Clemens’ paper is a literature review which is why it’s unsurprising he’s not alone: Hamilton and Whalley (1984) suggest a gain of 147%, Moses and Bjorn (2004) suggest 96.5%, Klein and Ventura (2007) suggest 121%.
Clemens’ review has not gone unchallenged and here’s where we meet Borjas again – he provides the only critique worth talking about. In his book Immigration Economics (2014) he argues that the costs of moving essentially wipes out the gains to world GDP. Borjas states that ‘“breakeven” cost of migration [is]… around $140,000’ and that ‘much of the presumed gains from open borders would disappear if migration costs were roughly of the same magnitude as the estimates in the literature’ (p.169, 262). It clearly doesn’t cost immigrants $140,000 to move across borders so what does Borjas mean? He explains:
In a world of income-maximizing agents, the stayers are signaling that there are substantial psychic costs to mobility, perhaps even on the order of hundreds of thousands of dollars per person, and that they are willing to leave substantial wage gains on the table (p.168).
You read that right. Borjas is saying the fact that some people stay means that its because there are “psychic costs” to moving. These costs are hundreds of thousands of dollars. I realise I’ve probably just repeated the extract above but I’m hoping you’re as incredulous as I am. Milton Friedman’s phrase about ‘looking how people vote with their feet’ comes to mind. What can we show by way of the empirical literature? Well, the first point is that Borjas is relying on the slightly ironic position that immigration controls don’t matter – people are choosing not to come. Clemens and Pritchett (2016) explain:
This claim implies that the existing global limits on international migration—passports, visa restrictions, limits on recognition of professional credentials, all deportations, all sea patrols, all fences—do not collectively have important effects on workers’ decisions about where to locate. This would be the case if, as Borjas asserts is possible, migration itself generally conveys sufficient disutility that those policy barriers do not substantially alter workers’ choice of location. This opinion is incompatible with existing evidence.
Second, we can compare wage differentials (i.e., the amount you could make if you moved to X-land from Y-land) in contexts where there are border controls and contexts where there aren’t – and we don’t need to guess about this as Clemens and Pritchett (2016) have looked into it. Their findings are, for me, the strongest economic argument for looser immigration controls. They find that when you have strong immigration controls, you get abnormally high wage differentials:
[Borjas’] Compensating Differential Case encounters further difficulty in explaining why there is a 600–800% real wage gap between Haiti and the United States (which are separated by tight visa restrictions and naval interdictions), but historically similar Guadeloupe exhibits only a 60% difference in real wages with metropolitan France (to which Guadeloupian workers may move at will). Similarly there is a 300% real wage gap between observably identical Filipinos in the United States and the in the Philippines [where restrictions are high], but only a 50% wage gap between ethnic Guamanians in the mainland United States and in Guam [where restrictions are less]
This discrepancy suggests that it’s not that there are “psychic costs” but that there are barriers to immigration that are stopping people.
Third, a point linked to the first, the preferences of people would likely change in response to changes in border controls so that when Borjas claims that people don’t move now, it doesn’t mean anything for what people would do if they had fewer restrictions. There is a large literature about ‘reference-dependent preferences’. The gist of these studies show that people value things they have over things they don’t because, for example, losing something you have (your homeland) is worse than losing something you don’t (life in the West) especially when you think it is improbable to get the gain (i.e., are you actually likely to get access to the West? If not, the loss aversion is high).
And this is precisely why Borjas is wrong: immigration restrictions lead to the same kind of loss aversion (which Borjas takes as evidence of “psychic costs”) that impact preferences (Koszegi and Rabin (2006)) and accordingly the “cost” is not as high as stated, certainly nowhere near wiping out the consistent gain found in the literature.
That said, it is right to have some scepticism over the higher estimates of world GDP especially because not everyone who is pro-immigration wants the elimination of all barriers, they may simply want high levels of immigration. Doquier et al (2015) find that the global GDP gains when you actually look at more realistic migration levels (and desires) is between 7 and 18% increase in global GDP. Table 3 below shows the income gains for workers in a country (including immigrants following liberalisation), natural (natives in the country receiving immigrants) and stayers (those who stay in their country of birth).
The Dcoquier et al study is, in my mind overly pessimistic, but it does two good things: first, it measures potential migrants and the educational composition of such migrants. Education in Fiji is worth less than education in the West (see endnote 7), so a skilled migrant with the same qualification as a native is not worth as much and their income gains wont be as high as some of the previous studies assume. But even on that basis, it is clear that immigration is good for the immigrants themselves and global GDP.  
This leaves three things to discuss. First, the brain drain; immigrants who come here may have better outcomes but this may have negative outcomes for those left behind. The logic of the argument is clear: immigrants are likely to be productive members, perhaps the most productive and when they migrate they deprive their home nations of their productive goodness. And it’s borne out by several studies in the literature – to take one of the recent examples, Ha et al (2016)[6] look at the migration that happens within China and find that ‘one permillage increase in the emigration rates will decrease the growth rate of that province by 0.4%.’
Here’s the interesting thing though: Ha et al (2016) find something else that it prevalent in the literature: migration increases the human capital of those who stay behind. Chands and Clemens (2008) find that a large exit of (skilled) Indians was associated with a large increase in human capital investment that human capital stocks rise, even net of large departures of skilled workers.’ Why is this the case? It’s because, rather sadly, watching people move out encourages people’s to want to do the same – and so they also go through the same steps that the migrant has but don’t end up leaving. This line of research has some interesting implications for the signalling model of education, please see [7] if you are at all interested.
Paul Collier (2013) in his book, Exodus, states the following:
For developing countries as a group the [brain drain] concern is clearly misplaced: gains outweigh the losses… [However] Intractable poverty as a problem.. is becoming concentrated in small, poor countries that have suffered significant net losses of their skilled population. As their diasporas build up, their rate of emigration is likely to increase. For these societies, “brain drain” unfortunately remains the right concern (p.203)
It’s clear that Collier’s statement is driven by Doquier and Rapoport (2012). Here are their findings, I’ve highlighted something which Collier neglects to (an unfortunate feature throughout his book):
On the whole, the simulations results reveal that the countries experiencing a positive net effect (the “winners”) generally combine low levels of human capital (below 5%) and low high-skill emigration rates (below 20%), whereas the “losers” are typically characterized by high high-skill migration rates and/or high enrollment rates in higher education. There appear to be more losers than winners, and the losers tend to lose relatively more than the winners gain.
The main "globalizers" (e.g., China, India, Indonesia, Brazil) all experience modest gains while many small and medium- size sub-Saharan Africa and Central American countries experience significant losses. However, the absolute gains of the winners exceed the absolute losses of the losers, resulting in an overall gain for the developing world as a whole.
I think this a neat finding but I’m not sure how much the ‘small, poor’ vs ‘developing nations’ dichotomy works. The aforementioned Chands and Clemens paper was based on Fiji and similar findings have been found in Cape Verdi (Batista et al (2011)), Papa New Guinea and Tonga (Gibson and McKenzie (2011)). It’s worth bearing in mind that the Doquier and Rapoport paper relies on not on real data but simulations using quite complicated algorithms and formulas that, to be quite honest, I don’t fully understand. They and Collier are probably nonetheless right that some countries suffer from brain drain whilst others do not – but I really don’t want to over-egg it because of that last line above. Beine et al (2003) find the same thing in their sample of 50 countries, they find
…the brain drain appears to have negative growth effects in countries where the migration rate of the highly educated is above 20% and/or where the proportion of people with higher education is above 5%. [By contrast] We found that most countries combining low levels of human capital and low emigration rates of their highly-educated are positively affected by the brain drain.
In that latter category of winners are 80 to 94% of the developing world (the former is from the Beine et al study, the latter is from Legrain (2002), p.181). And that’s without even taking into account returnees. There is evidence of massive human capital gains for the I.T. industry in Indian (Commander et al (2004)). But the returnees argument isn’t very strong, it’s just a slight mitigator. Why? Dustman and Weiss (2007) look at the UK and find that
….remigration for immigrants from Europe, the Americas and Australasia as well as the Middle East, other Asia and other countries is substantial (more than 45 per cent have returned after five years since arrival, compared with those who are still there after year 1), and seems to continue after five years, return migration for the other two groups is much less pronounced. There is little indication of any return for immigrants from Africa and the Indian Sub-Continent
The evidence from the U.S. is similar: Mayr and Peri (2008) find that Asian and European migrants have a return rate of around 20%  (or 0.8) but the Latin America return rate is close to 1 (both this study and the Dustmann and Weiss study use the same research design hence you can compare with Figure 3 above). The point is not that “don’t worry about immigration, a lot of them will leave” – but “the brain drain isn’t as bad because some people go back”. And what makes return migration a particularly good mitigator of the brain drain is that there is an emerging consensus those that do return are the mostly highly skilled (Mayr and Peri (2008), Dustmann and Weiss (2007), Gundel and Peters (2008)).
Another way to mitigate the effect of the brain drain brings me onto the second thing to discuss: the effect of remittances. di Giovanni et al (2014) has one of the best papers on the overall welfare gains (in terms of looking at market size effects, productivity gains and population effects and, importantly, remittances).

… the long-run impact of observed levels of migration is large and positive for the remaining natives of both the main sending countries and the main receiving ones. Relative to the counterfactual scenario in which no migration takes place, some countries in both groups are as much as 10% better off. Interestingly, while the overall numbers are similar, the salient reason for the welfare changes is different. For the countries with the highest immigration rates (Australia, New Zealand, Canada), migration raises welfare through increased equilibrium variety. For the countries with the highest emigration rates (El Salvador, Jamaica), the staying natives are better off because of remittances
Hidden in these averages are important nuances (which, again, shouldn’t be over-egged):
 
…. in the long run the large majority of OECD countries would be worse  in the absence of migration. The average OECD country would experience a welfare change of -2.38%, with substantial dispersion in outcomes (standard deviation of 3.07%). In this group, the largest losses are experienced by the natives of the countries with the largest observed shares of the foreign-born in the population: Australia (-11:63%), Canada (-7:07%), and New Zealand (-6:89%). However, it is worth noting that a handful of OECD countries would experience welfare gains: Greece, Korea, and Portugal would all be about 1.1-1.4% better off in the no-migration counterfactual
…. the majority of non-OECD countries also have lower welfare in the no-migration counterfactual, although dispersion in country outcomes is substantial. The average welfare change is -2.00% with an associated standard deviation of 3:55%. The highest welfare losses are to native stayers in El Salvador, the Dominican Republic, Jamaica, and the Philippines, at around -7-10%. Interestingly, a handful of non-OECD countries experience welfare gains: mainly, Trinidad and Tobago (5:70%), Mexico (1:32%), and Turkey (1:07%).
It is particularly interesting to compare the predictions for El Salvador and Trinidad and Tobago. These two countries would experience similar population gains due to return migration, at 19% and 17:9% respectively. But while the former would suffer a welfare loss of -8.72%, the latter would experience a welfare gain of 5.70%... the diverging effects of return migration on these two countries are explained by the role of remittances. 
These results are because of the fact that remittances help mitigate the effect of leaving populations – but not all countries get the same amount of remittances. The Table above shows the percent change in the real average income of natives of that country in the no migration scenario relative to the benchmark.
The effect, however,  of remittances also stands alone as an argument in favour of high levels of immigration. The literature on remittances is large and so I’m endeavouring to go through all of it. The graph above shows how remittances compare to other forms of transfers and flows. One thing that no one will question is the sums involved, see graph above (taken from Anghel et al in Handbook of the International Political Economy of Migration (2015), p.236).

Collier (2013) makes the laughable claim that ‘remittances largely offset the loss of output [from the immigrant leaving]’ (p.208). The difference is that there are now a few less mouths to feed and so per capita expenditure can be a little higher.’ This, of course, does not take into account the human capital effects explained above. But more importantly, the effect that his on poverty is phenomenal:
The World Bank has calculated what would happen to poor people’s incomes in a cross-section of thirty-seven countries if remittances dried up. In the countries where remittances account for a large share of the economy – 11 per cent of [GDP], on average – the share of the population living on less than a dollar a day would rise by half, from 24.8 per cent to 37 percent… Household surveys suggest that remittances have reduced the share of people living in poverty by 11 percentage points in Ugdana, 6 percentage points in Bangladesh and 5 percentage points in Ghana (Immigrants, Legrain (2007), p.164-165)).
Adams and Page (2005) looking at 71 countries find that ‘remittances do in fact, reduce the level, depth, and severity of poverty.’ Their estimate is that estimates that a 10 per cent increase in official remittances leads to a 3.5 per cent decline the share of people living in poverty. Jongwanich (2007) looking at Asia and Pacific countries find that a 10 per cent increase in remittances leads to a 2.8 per cent reduction in poverty incidence.
Ahmad et al (2010) finds that for Pakistan ‘the probability of households becoming poor decreases by 12.7% if they receive remittances’ and that the poverty headcount ratio declines by 7.8%. Yang and Choi (2007) have a neat research design using rainfall in the Philippines. Bad weather can cause dramatic decreases in income and remittances are found by Yang and Choi to offset 62.9% of this decline.  
And on and on the literature goes. In a perfect example of Collier’s back and forth, he does conclude his section on remittances by admitting remittances ‘have been beneficial and substantial for the people left behind in some of the poorest countries of origin.’ The picture on remittances isn’t all rosy in four respects. First, for those who care about inequality, as Anghel et al (2015) note the literature is quite mixed:
A large number of studies show that remittances have a negative impact on income ine-quality, as measured by the Gini coefficient (Barham and Boucher 1998;Rodriguez 1998; Adams et al. 2008b; Adams and Cuecuecha 2010a), and this effect is more pronounced when remittances come from international as opposed to internal migrants… However, De and Ratha (2005) found that the Gini coefficient drops for the effect of remittances; while McKenzie and Rapoport (2007) found that even though the initial effect of migration is to increase income inequality, as the level of migration increases, income inequality decreases. Given the mixed findings, the relation between remittances and income inequality is still a topic of debate (Handbook of the International Political Economy of Migration (2015), p.240)
Second, what are remittances actually spent on? My favourite study of the effect of remittances is Yang (2008). In 1998, the East Asia crisis meant that there was a devaluation of currency in some countries and the converse in others. This extra money (for countries who had their currencies devalued) went toward business investment and education. But again, the literature on this point is not as emphatic as the poverty-reducing effects, as Anghel et al note, there is a view that remittances ‘cause behavioural changes and are spent on consumption rather than investment goods’ (p.240).  
The third concern about remittances is that they may reduce labour force participation. To cut to the chase, the effect is likely around zero but the composition changes in response to remittances. Amuedo-Dorantes and Pozo (2006) show that Mexicans receiving remittances increase their work in the informal sector although there is some negative effect for women’s participation in the labour market in rural parts of Mexico.
The fourth concern is the effect on growth. Some studies show negative effects, some show positive effects, most show no effect (again, I can do not better than the summary in Anghel et al (2015)). My view goes the plurality of studies. I’m not particularly concerned about this because of the overall effects of migration on growth, trade and globalisations. Here are the results of a meta-analysis undertaken by Genc et al (2011):
…we analyzed the distribution of immigration elasticities of imports and exports across48 studies that yielded 300 estimates. The results confirm that immigration boosts trade, but theimpact is less on trade in homogeneous goods. An increase in the number of immigrants by 10percent increases the volume of trade by about 1.5 percent.
Why would this be the case? Well, migrants create networks in the destination country that have fewer costs to transacting with their origin state. Rauch and Trinidade (2002) find that the presence of a migrant Chinese population can increase in biltareral aid by 63% (their upper estimate is 102%). There’s also some evidence that the presence of migrants increases foreign direct investment into the origin country (Javorick et al (2011), for example, find that ‘a 1-percent increase in the migrant stock is associated with a 0.35–0.42% increase in the FDI stock’).
The aforementioned four concerns are worth talking about, but I’m not sure they would affect my calculus much in the face of the aforementioned poverty reduction effects. There is however one finding that bothers me and its one of the reasons I am for high levels of immigration but not open borders. How does the level of restrictiveness of immigration interact with the level of remittances sent? Doquier, Rapoport and Salomone (2012) look at this question.
Using a new database obtained by merging various second hand sources on bilateral remittances for a large set of country-pairs over the period 1985–2005.. [they find] The results strongly support the theoretical analysis, suggesting that immigration policies in the migrants' host countries determine whether the home countries receive relatively more or less remittances from their skilled emigrants
[The] more restrictive destinations are associated with skilled migrants sending relatively more remittances). The marginal effect of costly family reunion immigration policies on the propensity to send remittances by skilled people is estimated to be equal to 0.267… As expected, the sign of the coefficient of the interaction term is negative and highly significant. Our estimate of the marginal effect of skill biased immigration policies on the propensity to send remittances is equal to 0.79.
This makes sense: if immigration is less restricted, the ‘left behind’ are more likely to come over themselves. I can imagine a migrant thinking “why should I send money, when they can just come over?”. I think this is a devastating finding to those arguing for open borders. Nine times out ten when you hear people talk in favour of it, they’ll talk about remittances. But their system would likely destroy a significant amount of those gains! It’s also worth noting that what changes is the preferences of the migrants, rather than the number of people in need of remittances.
The final thing to consider, but only in a preliminary fashion, is the effect of national IQ on economic outcomes. Like I said, I am only looking at economic outcomes independent of institutions in this section. Garret Jones’ The Hive Mind has the following graph:
Jones doesn’t make this argument, but I will because it provides a good Segway to Part II of this series. National IQ has a positive correlation with GDP and GDP per capita. What if importing immigrants with lower IQ damages GDP? Well, the evidence on GDP and GDP per capita from the domestic section above would seem to refute this, but a more interesting argument is that even if it did affect GDP per capita, these are compositional effects. For example, if the average wage is 25k, and we let in a bunch of people whose average productivity / IQ is worth £16k, the average wage will drop. But the average for non-migrants is unlikely to have been changed.  
Jones does make an argument about the long term effect on institutions of immigrants. The following extract is a nice way to round up the literature above and conclude this part:
The economics of less-skilled immigration to richer, more productive countries are reasonably clear: life-changing good news for the immigrant with only fairly small effects one way or the other on so-called “native” less-skilled workers. That’s true when we look at the short run or when we look across towns and cities within the same country…
[However] the possible—I emphasize possible—effect on long-run institutions. Will less-skilled immigrants tend to vote for policies that will weaken the wealth-creating opportunities they’ve enjoyed? Or will less-skilled immigrants and their descendants instead build up high levels of human capital, perhaps raising the average information levels of voters? Conversely, might more skilled immigrants bring a focus on the long run, a more informed perspective, into political discussions? (p.161-162).
These are questions I’ll be looking at in the next part. Thanks for reading!
Endnotes
 [1] Eric Kauffman undertook an analysis of the BHPS and found that ‘there is no difference between pro-immigration whites and anti-immigration whites in their propensity to leave a diverse area.’ He goes onto conclude:
…local diversity does lead to more tolerant white attitudes and this is not the result of white flight. As more locales become diverse, this should lead to interethnic contact and more positive white attitudes to outgroups.
Kauffman, in an earlier post, noted the following:
We netted around 1700 white British respondents of whom about 200 said they had moved to a more or less diverse ward over the past decade. Whether the question asked about comfort with a boss of a different race or a Prime Minister of a different race, anti-immigration views or neighbourhood minority comfort thresholds, the result was the same. Namely, that racial and immigration attitudes had almost no effect on white mobility. Only at the conservative extremes did attitudes affect behaviour, but this was a marginal effect operating on 1 or 2 percent of the sample.
[2] I am not convinced that the predominant reason for the majority of people being anti-immigration is racism. Consider the fact that of the foreign born population, a majority is still for reducing immigration “a lot”. If you look at the graph below from Ipsos Mori, you can see that more recent immigrants agree with this anti-immigration position less.
Ipsos says that this may be a function of their time in the UK and the ‘type’ of immigration. It is more difficult for a recent immigration to say “put up the drawbridge” than a more established one – but in relation to the ‘type’ of immigrant’, the most persuasive confounder is age. I would like to see a breakdown of the foreign born population’s views broken down by age. In any event, the point is that given that a foreign population is anti-immigration, the argument that anti-immigration views are xenophobic or racist is more difficult to sustain. Admittedly, it might be possible that all the previous immigrants have non-racist reasons and the natives have racist reasons but I’m unconvinced by this given the downward trend of racism and the fact that only approximately 20-30% of people say they’re prejudiced themselves. Indeed, we merely need to look at why anti-immigration people give, from the same Ipsos poll:

They could all be liars, or maybe, just maybe, people can be reasonably opposed to immigration without it being racist despite it being wrong. The idea of being reasonable but wrong is an idea I’ve gone on and on about and I’m sure I will return to it at some point.

[3] On the subject of the impact of immigration on voting Leave, the main graph to take heed of it the one below provided by the Economist. They note that ‘The proportion of migrants may be relatively low in Leave strongholds such as Boston, Lincolnshire, but it has soared in a short period of time. High numbers of migrants don’t bother Britons; high rates of change do.’
 
http://cdn.static-economist.com/sites/default/files/imagecache/original-size/images/2016/07/articles/body/20160716_brc890_0.png
However, Colantone and Stanig (2016) (for more on this study see endnote 5) find that:
 
https://img.washingtonpost.com/blogs/monkey-cage/files/2016/07/STANIG-Figure2.png


…when London (which contains five regions) is included in the analysis, we find a negative correlation between arrivals and support for Leave — meaning that on average, regions with more new arrivals have lower support for Brexit. But once we remove those five regions from the analysis, there is really no pattern left in the data. This means that the correlation is driven by London, which is historically more cosmopolitan and diverse. For this reason we do not want to make much of this negative correlation — but are, at a minimum, confident that there is really no detectable relationship between how many immigrants arrived in recent years and how much support the Leave option received in the referendum.
 
[4] If you believe in rights, there is a very interesting meta-argument about whether we should give precedence to rights to self-determination or rights of association (which is, in essence, an application of loose immigration controls). I will return to this argument in a future part to this series. However, it’s worth thinking about what this means for mandate-era Palestine. If you believe that rights of association should trump the will of the people, then you must agree with British policy during 1917-1933, 1945-1947 of permitting Jewish immigration into Palestine. It’s likely that the local population, had they been in control of their country, would have voted to stop immigration.
 
My own view is that post-WW1, the British needed to provide stable institutions for the Ottoman Empire’s breakup. Indeed, the purpose of the Class A mandate was that they required “the rendering of administrative advice and assistance by a Mandatory until such time as they are able to stand alone”. World War II provides a later of justification for Israel not because of the Holocaust but because British rule of Palestine became doubly justified in attempting to avert Nazi rule of strategically important areas. Accordingly, during the period of legitimate British rule, the policy of open immigration (if that is indeed the objectively correct policy), was justified. When it then came to the establishment of states in a state-less area, precedence should be given to self-determination – and that is a liberal (relatively) non-nationalist justification for the State of Israel’s legitimate existence and establishment.
 
This gives moral justification for the existence State of Israel but it does not say anything about the creation of the refugee crisis between 1947-1948. During the 1980s following a series of books published by the New Historians, most notably Benny Morris’ The Birth of the Palestinian Refugee Problem, showed that refugee crisis was not borne about because, as Zionist historians claimed, they were ordered to leave by Arab commanders but through a combination of leaving through fear, direct involvement in the war and Israeli expulsion orders.* I’ve come to the view that the debate about the causes of 700,000 or so refugees doesn’t matter: whether they left voluntarily or not, post-1948, Israel stopped them from returning. There’s a discussion to be had about whether this was right, but it makes a discussion about the causes of the refugee crisis of academic interest and nothing more.
 
* The precise breakdown of the first wave is provided in Morris, ‘The Causes and Character of the Arab Exodus from Palestine: The Israel Defence Forces Intelligence Branch Analysis of June 1948’, Middle Eastern Studies (1986) Vol 22, Issue 1, 5 available at https://docs.google.com/file/d/0B-5-JeCa2Z7hMXpOaWotOTY0Wm8/edit
 
[5] The same argument could be made about globalisation. Take China’s move toward a market economy in the 1980s. The move toward a capitalistic model has lifted millions of people out of poverty. However, we know from the work of Autor et al (2016) that it has cost some people in the West. He finds that around 1 million of the 5.5 million jobs lost in manufacturing in the U.S. is down to the “China shock” – trading with China. Those whose jobs are displaced don’t find it easy to adjust: equilibration is ‘remarkably slow, with wages and labor-force participation rates remaining depressed and unemployment rates remaining elevated for at least a full decade after the China trade shock commences.’
 


However, as the above graph shows, we know that free trade and globalisation is a good – for those in China and the U.S (see, generally, Norberg (2003), Bhagwati (2007) and Bhagwait and Panagariya (2014). To take the most recent TPP deal, estimates of raising American incomes by 0.5% (on average) makes trade deals worth doing, TTIP will raise U.S. GDP by 3% (see here). These are not trivial gains. But most importantly, global utility is raised: China's poverty rate since it has been following more market-orientated has lifted millions of people from poverty.
 
The particular relevance of the globalisation discussion is a more recent study by Colantone and Stanig (2016). They summarise their findings in a Monkey Cage post:


Chinese import shock and support for the Leave option in the Brexit referendum, by region. Data: Eurostat Comext and Electoral Commission UK. Figure: Piero Stanig and Italo Colantone.


We found a strong and statistically significant relationship between the strength of the import shock and the Leave share in the referendum... Let’s look at Inner London, where 28 percent voted Leave, and Derbyshire/Nottinghamshire, where 56 percent did. That’s a difference of 28 percentage points in support for Brexit. According to the data and our statistical analysis, of these 28 points, at least 18 are attributable to the difference in the intensity of the “import shock” between the two regions.
 
[6]Incidentally, one of many of Mao’s devastating policies was the establishment of the hukou system which, essentially, branded people as agricultural or non-agricultural and stopped the movement of people who lived in rural areas to urban areas. As Ha et al note the ‘gross loss induced by the labor market segmentation from 1960 to 1978 amounted to 20%–60% of GDP.’ Commies are the worst.
 
[7] If you look at Docuqier’s model he says that one of the reasons the benefits of immigration is overstated is because we don’t take into account their education levels. It’s simply the case that, with some exceptions, most university educated immigrants come from universities that aren’t as good as Western universities. Accordingly, in Docquier’s model, he gives the following example: ‘each college graduate from Angola is considered as a combination of 0.73 of an actual college-educated worker and 0.27 of a less-educated worker.’
 
The issue is put more starkly by Mattoo et al (2008) which looks at the discrepancy across immigrant groups:
 
Even after we control for age, experience and education level, we find that highly educated immigrants from certain countries are less likely to obtain skilled jobs. For example, a hypothetical 34 year old Indian college graduate who arrived in 1994 has a 69% probability of obtaining a skilled job while the probability is only 24% for a Mexican immigrant of identical age, experience and education.
 
It’s interesting to consider is what this means for the signalling model of education debate. We have the following options as to explain this disparity:
  1. The market discriminates against immigrants irrationally / based on taste; or
  2. Signalling model is wrong: the immigrants are not as skilled despite having equivalent qualifications because their universities have not given them said skills; or
  3. Signalling model is correct: the universities aren’t as good and so the signal it sends to the market is “immigrants are not as good”.  
We can discount (1) assuming Becker’s model of discrimination holds true: the market punishes employers who discriminate. If you don’t want to hire someone on the basis of taste discrimination, you essentially impose a cost on yourself. I believe this model to be true, see for example Weber and Zulehner (2014) who look at discrimination against women and how this impacts the survivability of said companies. They find that there is a


strong indication for a negative effect of relative female share on exit rates [i.e., going out of business], which is not diminished by the inclusion of a rich set of other productivity relevant variables in the regression model. This effect is mainly concentrated at the bottom of the distribution: firms with relative female shares in the bottom quartile exit about 18 months earlier than firms with median share of females… highly discriminatory firms that manage to survive submit to market powers and increase their female workforce over time.
 
Levine et al (2012) is another excellent example. In the 1990s, there was a liberalisation of the U.S banking system. It may surprise many to know that nationwide banking was only permitted in from the 1990s. Levine looks at the effect of the deregulation in the 1990s on the disparity between black and white Americans. The deregulation had two impacts: first, it allowed the entry of non-financial firms which could expand credit and second, consistent with Becker’s model of taste discrimination, the market forced those with bigoted views to incur financial cost where they hadn’t before. The result was a rise in relative wage rates between blacks and whites:
The graph on the right shows the percentage change in relative wages of blacks in states where racism was effectively higher than the median and the right shows the percentage change in relative wages of blacks in states where was racism was below the median. Importantly, there is an increase in both thereby proving the benefits of increased competition and the validity of Becker’s model.  



So, what's true out of (2) and (3)? Annoyingly, I haven’t seen papers that directly address the question in a convincing way. Freidberg (2000) looks at the influx of immigration into Israel. As with the other studies, she finds the same discrepancy (see graph above). She also finds that 'the earnings gap between immigrants and natives can be fully explains by the lower value placed on the immigrants' human capital' and, indeed, once you account for this immigrants in Israel earn roughly 37% more than native Israelis. This doesn’t tell us much about the origins of the human capital though.

Mattoo et al (2008) mentioned above also look at this question. They look at “nominally identical” degrees among immigrant groups. They find that:

a large part of this country-level variation can be explained by certain country attributes. Some of these attributes affect the quality of human capital accumulated at home, such as expenditure on tertiary education and the use of English as a medium of education. Other attributes lead to a selection effect, i.e. variation in the abilities of migrants because they are drawn from different sections of the skill distribution of their home countries, and include the GDP per capita, the distance to the US, and the openness of US immigration policies to residents of a given country.
 
The fact that returns to education are impacted by the amount spent on tertiary education could be because under-funded schools don’t give people the skills they need or it could be that the market sees under-funded schools as not giving the right signals (or it could be both: lacks of skills -> no good market signal). Mattoo et al also find one other variable that matters: the effect of military conflict. As they go onto say:

the negative sign on the coefficient of the military conflict variable implies that the average quality of immigrants seem to increase with political stability… The existence of military conflict in the home country can have both a quality effect, because it weakens institutions that create human capital, and a selection effect, because it lowers the threshold quality of immigrants.
 
Again, military conflict leading to crapper schools makes sense but disentangling this effect from the signal that such education systems gives is difficult. Finally, I recently read The Knowledge Capital of Nations: Education and the Economics of Growth by Hanushek and Woessmann (2015). In one part of the book, they compare immigrants who have been taught in the U.S., vs those who have been taught in their countries of origin. Their main results show that being educated in the country or origin reduces average earnings in the U.S by 6 to 13% except for English speaking immigrants (where there is no reduction) after a series of controls. An increase in average test scores for those educated in their country of origin leads to 16% increase in earnings in the U.S. Hanushek and Woessman then go onto say:
 
The estimates do, however, provide direct support for the production view of schooling as contrasted with the signalling or screening view. As indicated above, one approach to identifying production versus signalling is to rely on what happens during schooling as opposed to the market returns to school attainment. The results above provide just such evidence because they show that the quality of different schools and the cognitive skills related to different schooling have direct payoffs within the same market (p100-101).  
 
I could be being incredibly dim but I’m not sure if this follows. It could be that the market is looking at signals for good schools – and this could account for the results. I do not have a view of whether the signalling model of education is correct, but I think future research designs utilising immigrants is an under-explored way of obtaining an answer.

 
Selected Bibliography

Adam and Page, 'Do International Migration and Remittances ReducePoverty in Developing Countries?', World Development (2005), Vol. 33, Issue No. 10, 1645 available at http://n.ereserve.fiu.edu/010040973-1.pdf
Ahmad et alm, 'Remittances and Household Welfare:A Case Study of Pakistan', Asia Development Bank Working Paper (2010) available at http://www.adb.org/sites/default/files/publication/28405/economics-wp194.pdf
Anderson, ‘Immigrants and Billion Dollar Startups’, National Foundation for American Policy (2016) available at http://nfap.com/wp-content/uploads/2016/03/Immigrants-and-Billion-Dollar-Startups.NFAP-Policy-Brief.March-2016.pdf
Amuedo-Dorantes and Pozo, 'Remittances as Insurance: Evidence from Mexican Immigrants', Journal of Population Economics (2006), Vol. 19, Issue No. 2, 227
Anghel et al, 'Migrant’s Remittances: Channeling Globalization' in Talani and McMahon (eds), Handbook of the International Political Economy of Migration, Edward Elgar Publishing (2015)
Aubrey et al, 'The welfare impact of global migration in OECD countries', Journal of International Economics (2016) Vol. 101, 1-21
Autor et al, ‘The China Shock: Learning from Labor Market Adjustment to Large Changes in Trade’, NBER Working Paper No. 21906
Beine et al, Brain Drain and LDCs’ Growth: Winners and Losers, IZA (2003) available at http://ftp.iza.org/dp819.pdf
Bitler and Hoynes, 'Immigrants, Welfare Reform, and the U.S. Safety Net', NBER Working Paper No. 17667 (2011)
Borjas, The Labor Demand Curve is Downward Sloping: Reexamining the Impact of Immigration on the Labor Market, NBER Working Paper No. 9755 (2003)
Borjas, Immigration Economics, Harvard University Press (2015)
Borjas, The Wage Impact of the Marielitos: A Reappraisal, Harvard University Working Paper (2015) available at https://www.hks.harvard.edu/fs/gborjas/publications/working%20papers/Mariel2015.pdf
Brauw and Russell, ‘Revisiting the Labor Demand Curve’ IFPRI Discussion Paper (December 2014), available at http://cdm15738.contentdm.oclc.org/utils/getfile/collection/p15738coll2/id/128872/filename/129083.pdf
Brian and Kovak, Immigrants Equilibrate Local Labor Markets: Evidence from the Great Recession, Applied Economics (2016) Vol.8, Issue 1, 257
Card, 'The Impact of the Mariel Boatlift on the Miami Labor Market', Industrial and Labor Relations Review (1990), Vol. 43, Issue No. 2, 245 available at http://davidcard.berkeley.edu/papers/mariel-impact.pdf
Choi and Salehyan, ‘No Good Deed Goes Unpunished: Refugees, Humanitarian Aid, and Terrorism’, Conflict Management and Peace Science (2013), Vol. 30, Issue No. 1, 53
Clemens ‘Economics and Emigration: Trillion-Dollar Bills on the Sidewalk?’, Journal of Economic Perspectives (2011), Vol. 25 Issue 3, 83
Clemens and Pritchet, 'The New Economic Case forMigration Restrictions: An Assessment', Center for Global Development Working Paper No. 423 (2016) available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2630295
Collier, Exodus, Puffin (2013)
Cortes and Pan, 'Foreign Nurse Importation to the United States and the Supply of Native Registered Nurses', Federal Reserve of Boston Working Paper (2014) available at www.bostonfed.org/economic/wp/wp2014/wp1407.pdf
Devlin et al, ‘Impacts of migration on UK native employment: An analytical review of the evidence’ Home Office and Department for Business, Skills and Innovation (2014) available at https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/287287/occ109.pdf
Di Tella and MacCulloch, 'The consequences of labor market flexibility: Panel evidence based on survey data', European Economic Review (2009) 49 1225
Doquier et al, ‘Efficiency Gains from Liberalizing Labor Mobility’, Scandavian Journal of Economics (2015), Vol. 117, Issue 2, 303 available at http://perso.uclouvain.be/frederic.docquier/filePDF/DMS_Liberalization.pdf
    - and Rapoport, ‘Globalization, Brain Drain and Development’, Journal of Economic Literature (2012), ungated version available at http://ftp.iza.org/dp5590.pdf
Dustmann and Weiss, 'Return Migration: Theory and Empirical Evidence from the UK', British Journal of Industrial Relations (2007), Vol.45, Issue 2, 236
Foged and Peri, 'Immigrants' Effect on Native Workers: New Analysis on Longitudinal Data', Centre for Research & Analysis of Migration (2015) available at www.cream-migration.org/publ_uploads/CDP_07_15.pdf
Genc et al, 'The impact of immigration on international trade: a meta-analysis', Norface Migration Discussion Paper (2011) available at http://www.norface-migration.org/publ_uploads/NDP_20_11.pdf
Ha et al, 'Brain drain, brain gain, and economic growth in China', China Economic Review (2016), Vol. 38, 322
Hanushek and Woessmann, The Knowledge Capital of Nations: Education and the Economics of Growth, MIT Press (2015)
Javorcik et al, 'Migrant networks and foreign direct investment', Journal of Development Economics (2011), Vol. 94, Issue 2, 231
Jones, The Hive Mind, Standford University Press (2015)
Jongwanich, 'Economic Growth and Poverty ion Developing Asia and the Pacific Countries', UNESCAP Working Paper (2007) available at http://www.unescap.org/sites/default/files/wp-07-01.pdf
Koszegi and Rabin, ‘A Model of Reference-Dependent Preferences’, Quarterly Journal of Economics (2006), Vol. 121, Issue 4, 1133
Levine et al, 'Bank Deregulation and Racial Inequality in America', Federal Reserve of Boston Bank Working Paper (2012) available at http://www.bostonfed.org/bankinfo/qau/wp/2012/qau1205.pdf
Legrain, Immigrants, Abacus (2008
Lisenkova et al, 'The Long Term Economic Impacts of Reducing Migration: the Case of the UK Migration Policy', NIESR (2013) available at www.niesr.ac.uk/blog/long-term-economic-impacts-reducing-migration#.VqNwz_mLTIV
Lucchino et al, 'Examining the relationship between immigration and unemployment using National Insurance Number registration data', NIESR (2012) Discussion Paper No. 386 available at www.niesr.ac.uk/sites/default/files/publications/090112_163827.pdf
Mattoo et al (2008), 'Brain waste? Educated immigrants in the US labor market ', Journal of Development Economics (2008), Vol. 87, Issue 2, 255
Markaki, ‘Sources of anti-immigration attitudes inthe United Kingdom: the impact ofpopulation, labour marketand skills context’, Institute for Social and Economic Research (2012) available at https://www.iser.essex.ac.uk/research/publications/working-papers/iser/2012-24.pdf
Mayr and Peri, 'Return Migration as Channel of Brain Gain', Return Migration as Channel of Brain Gain, CReAM (2008),  Discussion Paper No 04/08 available at http://discovery.ucl.ac.uk/14242/1/14242.pdf
Nickell and Saleheen, 'The impact of immigration on occupational wages: evidence from Britain', Bank of England Working Paper No.574 (2015) available at www.bankofengland.co.uk/research/Documents/workingpapers/2015/swp574.pdf
OECD, ‘International Migration Outlook’, OECD Publications (2015) available at http://ifuturo.org/documentacion/InternationalMigrationOutlook.pdf
Peri and Yasenov, 'The Labor Market Effects of a Refugee Wave: Applying the Synthetic Control Method to the Mariel Boatlift', NBER (2015) available at http://econ.sites.olt.ubc.ca/files/2016/03/pdf_paper_seminar_peri.pdf
Rauch and Trindade, 'Ethnic Chinese Networks in International Trade', The Review of Economics and Statistics (2002), Vol. 84, No. 1, 116 available at http://www.parisschoolofeconomics.eu/docs/koenig-pamina/article_rauchtrindade_ethnic-chinese-networks.pdf
Ruist, 'The fiscal cost of refugee immigration: the example of Sweden', Population and Development Review (2015), Vol. 41: 567 (summary available at www.voxeu.org/article/fiscal-cost-refugees-europe)
Wadsworth et al, ‘Brexit and the Impact of Immigration on the UK’, Centre for Economic Performance / London School of Economist (2016) available at http://cep.lse.ac.uk/pubs/download/brexit05.pdf
West, ‘The Diversity Delusion, Gibson Square (2013)
Yang, 'Remittances and Poverty in Migrants’ Home Areas:Evidence from the Philippines', (2005), available at https://www.cbd.int/financial/charity/philippines-remittance.pdf
-    and Choi, 'Are Remittances Insurance? Evidence from Rainfall Shocks in the Philippines', World Bank Economic Review (2007) Vol. 21, Issue 2, 219