How much human capital does Australia get via visas?

The Australian visa point-system is the envy of the world as it has ensured that Australia gets a large influx of well-educated, healthy, English-speaking migrants. How large is the free gift that comes walking into our doors this way? Conservatively, I would say 50 billion dollars per year, probably more. Let us go over the calculation.

One method of calculating the free gift of migration is to look at how much the roughly 180,000 migrants earn during their life in Australia. With a lifetime income in the range of 2-4 million dollars(not counting those migrants above 44 years of age or below 15), this would make the gift of migration in the order of 100-200 billion dollars a year, i.e. 10-20% of GDP every year. Yet, of course, this method is not valid because migrants also consume: they take up space, commute, drive up housing prices, etc. What they produce in their life is not the added benefit to Australia of their arrival because they will also consume part of what they produce. If there were perfect markets and no initial investments, then migrants would just get paid what they are worth and thus consume what they produce.

Yet, there are initial investments coming in with migrants, namely the cost of acquiring their human capital. It is that cost that Australia does not have to bear that is the net benefit to the community they came to augment. The value of a migrant is hence not really their lifetime income but rather the costs we do not have to cover because they walk in already trained instead of having to be raised and educated by means of local investments. In the simplest terms, the taxes on their income pays for the education and development of locals, but not for those already reared and educated  when they arrived.

One way to consider the gift we get is to calculate how much effort would be required to create the same outcome ourselves. What are the costs of educating and raising the same number of people to the same level as these migrants? Let us consider some estimates from the US and Australia.

The US department of Agriculture estimates that in America, it costs around 227,000 dollars to raise a child from birth till the age of 18. This includes food, shelter, the opportunity cost of the time investments of the parents, and school tuition fees. The cost of then getting that child through university is another 250,000 dollars, meaning that your average American university under-graduate represents and investment of about 477,000 dollars.

Now, in poorer countries these costs are lower whilst in some richer countries, these costs are higher. Yet, given that US GDP per capita is about as high as Australian GDP per capita, one could argue that 375,000 dollars will be close to what it costs in Australia to raise someone up to be a university undergraduate.

There are of course also estimates for Australia. A 2009 federal government report puts the costs at around 500,000 dollars per graduate , and Social Researcher Mark McCrindle estimates the costs to be as high as 1 million per 24-year old, but this includes costs of pregnancy and the supposed costs of children on accommodation (which is a bit much since pregnancy is partially consumption). Yet other estimates put this as lower, but still around 500,000 per undergraduate.

There are of course those who argue that people with kids are actually richer than most others and hence kids can’t be that expensive, but I prefer the studies that tabulate what actually goes into kids and to price those investments, which in turn reaches the half a million per graduate mark.

How many migrants do we get that fit this bill? Well, in 2010-2011 we got 113,725 migrants in the Skill Stream, and the planning level for the ensuing years is about 130,000. Yet, these are the number of people coming in under a particular Visa stream, which doesn’t just mean the skilled person but also his/her family and partner. Whilst the family and partner also already represent a significant investment of another country, the free gift is not equally as high as the skilled migrant themselves.

The productivity commission managed to dreg up the age-distribution of people in different visa categories. Presuming that their 2004-2005 distributions still holds, would mean that 66% of the skilled are in the 25-44 age category, and another 20% in the 10-24 years category, with basically no-one above 65. Indeed, at least 65% of the persons coming in under skilled migration have at least a bachelor’s degree. 16% have a postgraduate degree and the other 35% are mainly still at school.

As a rule of thumb, in the skilled visa category, we are getting around 100,000 migrants with a bachelor degree paid for by overseas countries. This represents a free gift per year of 50 billion Australian dollars, or 5% of GDP. If we would further count the well-educated in the family visa streams and count some of the education already in the kids who come with migrants, this is higher, but we could get lower numbers if we start counting the costs of the relatively less healthy and older in all visa categories. A more in-depth study would be needed to get a tighter number, but my ball-park guess is around 50 billion dollars per year, which makes immigration our most lucrative industry

Author: paulfrijters

Professor of Wellbeing and Economics at the London School of Economics, Centre for Economic Performance

16 thoughts on “How much human capital does Australia get via visas?”

  1. Nice exercise.
    Thinking in terms of the cost of education underestimates selection effects though. Let’s suppose for some reason there were 100,000 more babies born 25 years ago and we spent 500,000 thousand on them. (Being a fresh father I would first ask “man, who would change all those nappies”?!) Most of these newborns would be unlikely to be of university material while immigrants are already selected out in their home systems. So you wouldn’t get the same outcome you get via immigration.
    Also, you would need to acknowledge that the children (and grandchildren) of immigrants are also more likely to be of higher productivity (than an average Australian) due to the inheritance of this selection (not just due to genes but better parenting/role modelling).


  2. Well I am confused. 180,000 migrants earning say 40,000 per year means 7.2 billion – not 90-100 billion – per year. They pay tax on this and consume services. The difference between the tax they pay and the services they consume for an average Australian over their life time would not be that much I reckon. How much of your lifetime income do you die with? Not that much. This whole calculation doesn’t make much sense to me.

    The cost of education calculation is easier to think about. Economist should always be comparing one policy with another. If the choice is between one educated migrant aged 21 and one new-born baby, we are k$500 better off with the migrant. I would point out though that the logical consequence of this argument is that we would have a society with no children at all and perhaps 400,000 migrants to achieve a sustainable population.

    The choice however is not between a child and a migrant. People will have as many children as they desire.It kind of one of those basic human rights that folks insist on. It seems unlikely that the fertility rate will drop in Australia from the current level of about 1.8-1.9. Our population is close to sustainable with zero immigration. So, it seems to me the choice is between a migrant and no migrant.

    Or you could equivalently say that the k$500 the immigrant brings does not get given to us. It allows the immigrant to produce a few million dollars in their lifetime, most of which they will consume.

    So I think you are back to the first calculation.

    Surely the reasons for immigration are (i) to fill holes in the labour market e.g. doctors and plumbers, (ii) widen the perspective of the society. The dividend is efficiency. And I think you are right that Australia does this aspect of immigration better than others.


  3. Kenan,

    of course there are all kinds of other things. Selectivity indeed matters, but including that needs all kinds of additional assumptions including judgements on the ability differential between the migrants and the incumbents. The sort of thing a proper paper should address!


    the issue with your calculation is that at any point in time you have way more than 180,000 people who came as migrants: they stay more than a year. So it is an issue of not just counting the first year of their stay but the lifetime value of their stay, which gets you a much higher number than what they make in the first year of their arrival. Even more important perchance though is that one should not see their wage as what they add to the incumbents: it is their productivity minus their consumption that is the rent the incumbents get. So I think the confusion stems from the fact that we are talking cross-purposes. The 50 billion is the benefit to the incumbents of this year’s migrant intake. Their lifetime wages is another number and the impact on the overall economy of their arrival is yet another number.


  4. First, thinking statically is a killer. You need to think about immigration as not the flow of a person, but of a lineage entering the system at a given point. If they are highly educated 20/30 somethings, that’s great. But they will have children who cost the existing members of society to educate, medicate etc. They will also get older, and cost society in terms of their own health care and so forth. Even the PC models shows that the age-dependency ratio get worst with higher immigration.

    Second, a lot of skill that skilled immigrants arrive with are actually job experience. This means that someone looking to climb the corporate ladder will get crowded out by these people. Highly educated people need capital to work with. No good having a well trained engineer using a pick and shovel, that any able-bodied person could handle. So it really appears like owners of capital get to invest in capital without paying to invest in training people to use it (roughly speaking).

    Third, Australians mostly pay for their own university education through HECS. And if we keep importing university educated people, it will decrease the return to domestic university education – a bad outcome for existing youngsters.

    Forth, in the short term there are capital costs paid by existing residents to duplicate infrastructure for a larger population – roads, dams, pipes, parks, etc.

    Fifth, the distribution of benefits is important. As Chris suggests, if they earn even $100,000 each a year, that’s $10billion in wages. If they are really contributing $50billion to output (which the aren’t in your analysis anyway, although people will read it that way), who captures the difference? I would say mostly this return is captured in asset prices and corporate profits, rather than distributed amongst wages.

    Sixth, and probably most importantly (I hope everyone is reading this far), your model simply says – calculate the ‘cost saved’ of raising an educated person, and multiply it by the number of educated immigrants ($500k a pop). Which is obviously nonsense because it compares on scenario against a fictitious other – the same rate of population growth occurring through births. A better way is to compare is the total output with 100,000 skilled immigrants divided by (population+100,000) against total output without divided by population. Of course, this is another static model. We should also consider the long run demographic changes and the likely costs/benefits under the two population growth/education scenarios.

    So if you can’t tell, I disagree.

    In fact, I would almost go so far as to argue that having unskilled immigrants is better for the existing population, since it crowds out unskilled jobs making the return on education better.


  5. this is not a standard way of measuring the gains from migration which is normally to measure the benefits preexisting people get from newcomers. This is a measure net of gains to the newcomers themselves. If the newcomers are skilled they might boost demands for complementary unskilled labour. The substitution effects on demands for substitute skilled workers will be negative but the demand for locally-owned non labour inputs will be boosted – higher profits and rents to inputs locally owned. The net gains would be much less than the proprtionate growth in the workforce – much less than 1 per cent of gdp and thats ignoring costs of extra external diseconomies.


  6. to clarify: what is being measured is the cost of otherwise building up the human capital that comes walking in for free. So if Australia would want (for whatever reason) the same amount of human capital,the current people living here would have to invest 50 billion to build it up themselves. It is one measure of the gains to Australia. In a model in which human capital is build up via taxation on the workers, it is furthermore a measure of the cross-subsidy of migrants since they will be taxed the same (and more because for some time their entitlements are less) but their own human capital is already paid for by overseas entities.
    Yes, there are many other ways of measuring. The production function approach that Harry seems to favour is not my favourite because I think the issue of build up of human capital is more important. However, as I said, it’s a ballpark figure.


  7. If young Paul invests $100000 in a PhD in Holland and lives in Holland my welfare as an Aussie isn’t affected. If not so young Paul moves to Australia My welfare is likewise unaffected unless he participates in labour markets and influences other input markets or imposes externalities on me. These are the effects I trace out in my earlier comment.

    Most of the returns on Paul’s investment will accrue to Paul not to those who were here prior to Paul arriving. Not that Paul is therefore unwelcome – plausibly the small positive benefits will outweigh any negative externalities and negative effects on real wages.


  8. presumably the small positive benefits to existing Australians can be boosted if we get even more targeted about any financial/business capital the new Australian brings

    Australia is not only short educated/skilled human capital – in aggregate, we have a current account deficit that gets funded by selling assets to non-residents and making dividents and interest payments to foreigners.

    instead of offering permanent residency and all the associated rights, there are probably other pathways to Australian residency – leasing, securitisation etc. – if that is the goal of the person with the capital – human or financial


  9. If young Harry’s education in England cost England half a million, and he then comes fully educated here in Australia, spends his life working here, is taxed half a million more than he gets in services with the surplus spent on the next generation of educated homegrown Harry’s, Australian incumbents just indeed got half a million in welfare surplus from him walking in the door for free. Plus all the benefits of Harry providing skills in short supply here (some production complementarity between Harry and the incumbents), minus the cost via increased congestion, plus perhaps his already half educated kids, minus perhaps more needy relatives, etc, etc.

    To conflate the FDI that walks in with the added welfare toincumbents would only hold under very specific assumptions and they of course do not fully apply. Many of the expenses are not made via taxation and migrants do not get the same benefits from taxation as others, etc.

    To calculate properly one would need to have a life-cycle model and data of migrants which we don’t have as far as I know, partially because we lack the longitudinal data and detailed tax-welfare data specific to migrants to do it. The department of immigration stops following people after two years when they become residents so one cannot use that one. The Hilda has more information but lacks the data to work on the production side, etc.

    Care to try to calculate it properly, Harry? What I have seen so far misses all th tax and welfare stuff crucial to my argument, but perhaps I have missed a part of the literature. Till someone shows me a better calculation, the 50 billion FDI walking in is my best measure of the benefits of our skilled migration policy!


  10. Paul. I assumed that you would be thinking at the margin and trying to measure the value of 180,000. In any case, both you and all commenters seem to agree that the 100-200 billion figure is nonsense. And you are not even looking at output per head.

    “Till someone shows me a better calculation, the 50 billion FDI walking in is my best measure of the benefits of our skilled migration policy.” It is not an estimate of that at all Paul. As your title states, it is a measure of the human capital Australia “gets” via visas. The word “gets” here means that the capital resides in Australia. You and I do not get it.

    You ignored my comment about comparing one policy with another. If you are trying to value immigration, then I think is makes sense to compare wealth with and without the immigrant.

    Suppose an immigrant came and stayed in your house Paul. He pays rent, but you were intending to rent out the spare room anyway. He has k$500 of education and makes a good wage. How much better off are you? Sweet FA. He and his fellow migrants might inflate rents for you, but that doesn’t benefit everybody.

    Or are you saying that you would be better off having the immigrant boarder than your expensive kids who does not even pay rent? Then you are indeed k$500 better off because you do not have to educate a kid. This is the only way the k$500 can be accounted to you.

    Ideally, we want to follow the cohort of residents in some base year (say 2000) and compare the GDP per head of these people now compare with a prediction of what their GPD would be with some lower/higher value of NOM. There would be all sorts of second order effects that would determine the outcome. I think the first order effects are zero. Plug the numbers into some of the big macro model you trust and see what you get . But you will still have to factor in congestion and infrastructure.

    In the mean time, how about we start charging market rate for all visas? Now that would truly be an added value to residents.


  11. Chris,

    This was what my reply above with the fictitious Harry was about: if th new migrant in your neighbourhood has an expensive education paid for by some overseas government, you get him in the period that he pays more taxes than that he uses local education services, effectively meaning that via the taxman he is educating your kid for you.
    Congestion and other factors are clearly also important, but the key thing one would want in a more proper investigation is to follow the migrant over the life-cycle to the degree that you know their taxes and usage of government services. The method in the post is just a first glimpse at what you would end up in the situation that all the education investments are paid by taxes and all the other factors cancel out. Clearly not all education is paid for by taxes but clearly migrant initially use less services because of their lower rights, plus they bring in money, but they come with kids, etc., etc.


    1. I see the govt has introduced a new visa consistent with my earlier blog The investor visa Put up $5m, visit a certain no of days etc. older educated/skilled/wealthy. Assists with human and financial capital shortage


Comments are closed.

%d bloggers like this: