This is the draft paper on skilled migration I will give at the Australian Conference of Economist Conference in Perth next Monday. Even at this late stage comments are welcome.
1. Introduction. The Productivity Commission’s Position Paper Economic Impacts of Migration and Population Growth considers the effects of recent policy trends that place a greater emphasis on seeking skills in Australia’s migration intakes – about 70% of the intake was skilled in 2005/06 compared to only 29% 10 years ago.
There are enhanced demands for skilled migrants in Australia as a consequence of the Australian economy’s strong macroeconomic performance over the past 16 years and because demands for new skill mixes have emerged in the economy. Given that annual immigration is a small fraction of the workforce, and that migrants are not that different from most native-born Australians, one would not expect large effects from low levels of additional skilled migration and this is what the report finds. Using an economic framework, based on an open economy Swan-Solow growth model, the report finds that, maintaining a 50% higher skilled migration intake over 20 years – this would make total population 2.7% bigger than it would otherwise be – leads to workforce participation rising by 1%, average hours worked rising by 1.3% and per capita incomes rising 0.6% or by about $335.
Overall unemployment initially increases but ultimately falls a little. Apparently surprised by the small size of the gains here, the Department of Immigration and Multicultural Affairs commissioned Econtech (2006) to recompute these effects using their own Migration Modeling Framework. This latter study came up with higher gains – around 1.1 per cent rather than 0.6% in per capita incomes. The larger effect occurs because Econtech allow for positive effects on labour productivity of having skilled migration and for more rapid adjustment to pre-immigrant entry capital-labour ratios as the labour force grows.
There are no microeconomic foundations to either the PC or Econtech studies, a fact that obscures important core issues. The main effect identified in the PC report is the increase in labour force participation with implied effects of (i) capital dilution which reduces capital productivity and (ii) negative effects on the terms of trade that stem from the fact that, with more migration, the country imports more thereby offsetting favorable workforce participation effects. Neither of these effects seems particularly sensible as a way of understanding migration outcomes. The theory used is at best rudimentary and mechanical. There needs to be a focus on the microeconomic effects of having more skilled workers.
Skilled migrant workers provide gains to a destination economy through standard gains-from-trade effects. Berry and Soligo (1969) illustrate the ‘gains-from-trade’ effects in labour markets of labour migrations – a migration is equivalent to a market liberalization which will hence confer efficiency gains. The mechanism by which this occurs is a reduction in the marginal productivity of labour and an associated wage reduction that brings about a proportionately greater increase in returns to non-labour productive inputs. If there is wage stickiness the gains from labour migration will be reduced by increased unemployment of either migrants or residents so that implied unemployment levels are also an issue. Hence the extent to which an economy can access these gains depends on the flexibility and competitiveness of its labour markets.
With skilled migration there are further benefits to a destination country from implied skill externalities. While it is easy to capture these latter effects in a theoretical model I know of no satisfactory way of empirically estimating them. The answer is therefore to rely on sound theory not to conduct empirical exercises which ignore such effects or engage in numerical simulations to forecast effects based on guessed estimates of what are intrinsically hard-to-determine parameters.
The reason for focusing on skill externalities can be clarified by considering the implications of skilled versus unskilled migration in a general equilibrium model of market clearing without externalities. Here, if gains-from-labour-market liberalization are supposed proportional to differences in human capital endowments – so most gains occur from having the very skilled transact with the very unskilled - a premium will be attached to highly unskilled migrants. That this is counterintuitive suggests that the real focus of concern is on situations where highly-skilled workers get paid less than their social marginal product because of the external benefits they convey.
Finally, if we do allow for wage stickiness and resulting unemployment it is natural too to assess intakes in terms of their impact on unemployment in different skill categories. Unemployment tends to be concentrated among the unskilled so on this basis alone it makes sense to target skills.
2. Reformulation. Let me sketch what I think is a more sensible and robust qualitative analysis of the case for pursuing a skilled migration program.Having skilled migrants come to Australia implies a ‘brain gain’ - human capital acquired in other countries becomes available in Australia – this is the widely-discussed brain drain from developing countries: See Centre for Global Development (2005), World Bank (2005). Indeed, from a global viewpoint this gain is a profound ethical argument against skilled migration – the World Bank (2005) find that between 25-50% of college educated citizens of poor countries live abroad in an OECD country. Australia is a developed country which captures much developing country human capital without paying the full investment cost.
Ignoring the negative welfare implications of this important issue – as the PC report does – the arrival of skilled migrants would not represent as great a source of net advantage to resident Australians if arrivals were paid their marginal products and thereby captured the labour market gains created, leaving gains to the local economy only from effects on factor returns in other input markets. But skilled migrants do not get paid their marginal products – skills can be copied and ‘leak’ into the community – to take an extreme example, a brilliant scientist or a creative artist is paid less than his or her marginal worth with some value being transferred to owners of assets locally and complementary workers. If these owners and the complementary workers are residents of destination countries then gains are transferred to them. This benefit is augmented by progressive taxes which also transfer wealth on balance to residents.
Other advantages of having skilled migrants include input complementarities, or pecuniary externalities, with unskilled workers - having more skilled workers boosts demands for those less-skilled. Having more imported specialist doctors increases the demand for less skilled occupations such as receptionists, nurses and those in the service sector generally. Having more skilled migrants reduces pressures on the public purse directly because those with skills are less likely to be unemployed and, indirectly, by increasing the demand for those with lower skills who are otherwise exposed to higher unemployment risk.
In addition, selecting migrants with skills that are in short supply will in itself reduce any skill-mismatch component of unemployment. Skill mismatches in part reflect distortions in domestic labour markets that result from restrictions on resident entry into sought-after professions such as medicine and restrictions on labour market mobility through relocation costs such as stamp duty levies on house purchases. The first-best policy is to abolish these restrictions – and this is a more satisfactory policy in the first instance. But, if this cannot be done, directly targeting migrants with particular skills may outperform labour market retraining programs domestically even if migrants are imperfect substitutes for residents with the same skills. Having ‘footloose’ skilled labour migrants can compensate for immobile of skilled labour locally.
Finally, there is an interaction between the need to secure certain skill requirements in the population and population aging. The number of older workers in the economy is increasing while numbers of younger workers are falling or growing slowly. If labour shortages provide a stimulus to technological development and to higher productivity resulting from increases in capital per worker, this is not a problem provided older workers are substitutes for young workers. But in jobs that require the most sophisticated technological skills – what McDonald et al (2006) call complex problem solvers or CPS, older workers are not good substitutes for young workers. Psychology and economics show that complex problem solving skills deteriorate rapidly after age 40 and, consistent with this, in Australia, 80% of CPS are aged less than 40. Migration is a highly effective way of increasing the supply of CPS workers when the migration program is oriented towards highly-skill, younger migrants.
Generally favoring skilled rather than unskilled migrants has both favorable efficiency and equity implications. Having more skills provides an appropriate response to market signals which suggest higher returns to those with skills and this increases overall productivity. But, in addition, effects on the public purse and on the job opportunities for those less skilled, suggest that having a strong skilled immigration component outperforms having migrants with lower skills.Indeed while trade liberalization with low wage countries harms the unskilled by adversely impacting on their wages and working conditions, attracting skilled migrants can help this same group through pecuniary externalities or by at least concentrating unfavorable distributional impacts among those best able to afford it.
Ignoring the beggar-thy-neighbor issues that there is a strong case for pursuing an expanded skilled migration program if this can be done. If implemented correctly there will be low resulting unemployment effects, perhaps even a positive role in reducing unemployment of those less skilled, low adverse distributional effects and we will benefit from above-average gains-from-trade benefits by economizing on training costs and through the skill externalities that augment standard gains-from-trade benefits we get from additional migration. The key issue is: Is such a skills policy feasible?
3. Supply-Side Constraints. With low Australian unemployment skilled migrant entry is a substitute for training domestic workers to fill skilled positions. Large numbers of accountants and IT specialists are, for example, being sought in the subcontinent. These people are sought directly as migrants here or indirectly by encouraging untrained students to enroll in universities here as full-fee paying students who later apply for residency visas on the basis of their skills.
With respect to seeking skilled migrants directly an issue, neglected by the PC report, is whether or not there is a sufficiently elastic supply of sought-after skilled migrants who will fit into current skill niches at going wage rates. The assumption of perfect elasticity has never been particularly realistic (Clarke (1994, p.69-70)) but is less so today than in the past even though employment prospects in Australia are very good. It is not only Australia which offers good employment prospects to skilled workers from the developing world. Australia is competing with many other countries in the Western world (the US, Canada and France, for example) in seeking skilled migrants. All these countries have adopted ‘cherry-picking’ policies that select intakes on the basis of needed skills.
Indeed, because Australia’s migration intakes are a small proportion of total global intakes the supply of skilled migrants to Australia can be understood as partly an endogenously determined residual given the migration policies of other countries. There is nothing new in this perspective – I have argued elsewhere that this has been the sensible way to thing of the Australian migration program for much of the post-1860 period (Clarke (1994)).
It could well be that skilled migrants may assess employment and salary prospects in different countries net of resettlement costs with such costs being negatively related to the size of communities with the same ethnicity, religion or prior nationality already living in Australia. Indeed many of those with skills who migrate to Australia do have family linkages in Australia – a factor that reflects Australia’s competitive advantage in these situations. As levels of family-related migration decrease it is likely that this source of advantage will diminish.
It might be argued that maintaining a skilled-only migration system would eventually make us uncompetitive since we would not then be able to attract the same quality skilled migrant as Canada and the US in particular.An aspect therefore of attracting the best skilled migrant is the extent of culture and diversity in Australia. A skilled Greek (and other) migrant would find moving to Australia very attractive because it has a large Greek population with well-established Greek-oriented social institutions such as churches and schools. These were built by migrants who came here to form resident communities. Thus family and community migration make Australia more competitive in the market for skilled migrants. Family and community migration has been successful in Australia. Greek, Italian and Asian communities are doing well and have enriched our culture.
As an issue of practical policy design the best immigration policy is to have a combination of skilled and family migration. The latter might have a random component that seeks out a portfolio of ethnicities and community types that form a basis for attracting future skilled migrants. This will help us remain competitive in the market for skilled migrants.
4. Unsound claims in the PC Report. There are several features of the PC report which seem conceptually flawed and which limit the ability to discern the effects of augmenting the skill mix.
The claimed capital dilution effects of having migrants here and the effects of increased immigrants on our terms of trade seem erroneous to me. Capital stocks are not ‘diluted’ when immigrants come here because we don’t live in a communist state. You don’t acquire capital assets by walking through the Australian door – you have to acquire them by saving or bring them with you. Having extra skilled migrants here increases the productivity of capital at the margin in Australia which, in a world of mobile capital, will encourage a healthy current account deficit reflecting implied capital inflows. Having migrants here who demand more of our non-internationally traded goods mean expanded markets for Australian producers of these goods which imply welfare gains to residents that are analogous to trade liberalization. If immigrants buy cars and other consumer durables from overseas by spending their income that is in no sense a cost to resident Australians if these same migrants more than proportionately increase our capacity to export. All these effects are, in any event, likely to be very small given the limited annual migration intake.
In the PC report (and to a less extent in that of Econtech) the lagged adjustments in capital stocks to increased skill availability are implausibly slow and stretch even into the decades rather than years. That these sorts of forecasts are entertained seriously at all suggests the need for a rethink of microeconomic foundations. Current account deficits are strongly contemporaneously (and innocuously) linked to labour migrations – I cannot believe that it takes decades to augment sought physical capital stocks to accommodate new migrants.
Nor is the focus in the report on income per head a reasonable way of assessing the economic impact of immigration since this is defined inclusive of newcomers - it can fall with immigration though every group can be better-off as has been argued countless times over the past 20 years: See Clarke (1997) for a ‘recent’ attempt to clarify this elementary issue. The relevant target variable is the effect on incomes of non-immigrants. If this increases then everyone will be better-off (assuming voluntary immigration) even if overall income per head falls.
Moreover it is this misconception that provides the initial result in both studies that in per capita terms Australia will be worse-off short-term from increased migration. In my view this is a misleading implication of the modeling since both residents and migrants will be, on balance, better-off.
4. Other benefits from skilled migration intakes. It seems absurd to try to argue a case for permanent migration based purely on meeting short-term skill mismatches in Australian labour markets. Admitting intelligent, skilled people who are likely to be self-reliant and who will contribute to Australian society in broader terms is likely to be a more compelling basis for permanent migration policy. As residents of a multicultural society we derive social benefits from living in a stimulating cultural mix and having the opportunity to interact socially with people of different backgrounds. Economics is not the only basis – or even the primary basis - for assessing migration policies. But in many cases selecting people with skills on the basis of non-economic criteria might, in any event, advance economic objectives. Consider for example Australia’s environmental objectives.
The PC report does not really talk much about environmental impacts from un-priced congestion and pollution. The costs of these externalities are a problem with or without immigration but are worsened by it. The answer of course is to internalize these externalities using pricing or other policies thereby addressing these issues. It probably does not make a lot of difference between selecting skilled rather than unskilled immigrants to address such issues but, to the extent that equity arguments are used to thwart efficient pricing – it is difficult to congestion price roads because poor workers will be disadvantaged – there is a slight preference for selecting those with skills.
There is a case for accepting temporary skilled migrants for meeting short-term labour market imbalances. While market wage rates should provide clearer signals for desired training programs than do publicly-designed ‘job training programs’, such market processes operate with a lag and the skill-mixes sought may only be in temporarily high demand. In addition, market wages for those being trained are distorted by an unbalanced wage structure. Apprentices, for example, are often diverted into unskilled work with better-pay.
In dealing with fluctuations in demand for certain skills more flexible work and training regimes are sought. Temporary skilled migration programs have strength here in avoiding these constraints and provide greater labour market flexibility in coping with such things as the current resources boom. Government training programs and business programs, such as apprenticeships, come into their own longer-term but short-term temporary skilled labour needs can best be met by employing skilled migrants on a temporary basis. 5. Conclusions and
Final Remarks. Ignoring the brain drain implications of attracting skilled migrants from developing countries there is a strong, selfish case for focusing on skilled migrants in the migration program if this is possible.
Such a focus provides equity and efficiency gains to Australia, and is likely to reduce unemployment. The benefits are the traditional ‘gains-from-trade’ one gets from immigration augmented by skill externalities and favourable effects in job-markets from avoiding skilled mis-match. One also can expect better distributional and employment outcomes with a bias towards skills.
The main limitations in such an argument stem from the feasibility of such a policy. There are not many $50 bills lying on pavements. Unexploited gains from attracting large amounts of human capital without paying for it are low given the high competition in seeking skilled migrants from the US, Canada and large European economies such as France. The best chance of retaining a competitive edge in seeking such migrants is to community build in particular areas of the intake by retaining strong, attractive, family-oriented programs as well as those seeking skills. A diverse portfolio of intakes increases the opportunity to attract future sought5 migrants from many countries.
Although these conclusions above are couched in terms of economic benefits there are social and other benefits from gaining those with skills. Skilled intelligent people contribute to our society in many ways other than economic and the opportunity to interact with them is an unpaid-for benefit of emphasizing the case for skilled migration.
Tying the migration program mainly to acquisition of sought-after skills that provide broad community benefits without significant distributional costs also makes the migration program more politically sustainable. The ethnic vote-buying, predominantly family-based, programs of the Hawke Labor government were not sustainable in an Australian context – as capably discussed in Gruen and Grattan (1993). Debates still continue in Australia concerning the appropriate extent of diversity sought in the migrant intake and whether we do migrants with undemocratic, misogynist values and actively hostile attitudes to the Australian way of life. I don’t. Selecting immigrants who are educated and who have skills limits the need to even consider entry demands from such people.
R.A. Berry & R. Soligo, ‘Some Welfare Aspects of International Migration’, Journal of Political Economy, 77, 1969, 778-794.
B. Birrell, V. Rapson & T. F. Smith, Australia's Net Gains from International Skilled Movement, http://www.immi.gov.au/media/publications/pdf/aus_net_gais_int_kills_mnt_2004_05_.pdf, May 2006.
H. Clarke, ‘Labour Migrations and the Pseudoconvergence of National Living Standards’, Economic Record, June 1997, 73, 120-124.
H. Clarke, The Rationale for Forward Planning and Stability in the Migration Program, Bureau of Immigration and Population Research, Melbourne, 1994.
Econtech, The Economic Impacts of Migration: A Comparison of Two Approaches, Report for the Department of Immigration and Multicultural Affairs, Canberra, April 2006.
F. Gruen & M. Grattan, Managing Government – Labor’s Achievements and Failures, Longman Cheshire, 1993.
P. McDonald & J. Temple, Immigration and the Supply of Complex Problem Solvers in the Australian Economy, http://www.immi.gov.au/media/publications/pdf/immigr_supplyof_%20complexproblems_solversCPSDIMA.pdf, 2006.
Productivity Commission, Economic Impacts of Migration and Population Growth, Productivity Commission Position Paper, AGPS, Canberra, January 2006.
World Bank, International Remittances, Migration and the Brain Drain, World Bank, Washington, October 2005.
Centre for Global Development, The Global Migration of Talent: What Does it Mean for Developing Countries?, Centre for Global Development, October 2005
Reference:  Bob Birrell et al (2006) provide a detailed verification of the claim that Australia has been a very significant source of ‘brain gain’ up to 2004/05. This is due to the large numbers of skilled migrants arriving in Australia and to the return home of skilled expats.