The Treasury

Global Navigation

Personal tools

Treasury
Publication

Migration and Macroeconomic Performance in New Zealand: Theory and Evidence

4 The case for large negative effects

This section considers the possibility that migration could lead to large negative effects, either directly on labour or housing markets, or indirectly by diverting resources to less productive activities, or via a “brain drain”. Overall, the evidence points to small domestic labour market effects (which may be positive or negative), moderate negative effects on the housing market (which partly reflect other features of the New Zealand market that inhibit the response of housing supply to increased demand), and modest net costs from replacing skilled emigrants with skilled immigrants. Although this is difficult to establish empirically, it is also possible that meeting the infrastructure needs of migrants may divert resources from more productive activities, imposing potentially significant macroeconomic costs.

4.1  Labour market impacts

The impacts of immigration on the host labour market depend on the size of migrant flows relative to the overall population, differences in characteristics of immigrants and local workers, the overall labour market structure and adjustment processes, and the stage of the economic cycle when immigrants arrive.

Since these impacts have changed over time with technology, institutions and in response to immigration itself, this section looks at historical and contemporary impacts separately.

4.1.1  Historical labour market impacts

While bringing in additional labour when it is relatively scarce greatly increases wealth creation opportunities (see the earlier discussion on large positive effects of factor price equalisation in section 3.1), it also has distributional consequences. Hatton and Williamson (2006) argue that examining the period of migration before quotas and other policy barriers were applied can provide clearer insights into the impact of immigration.[106]

The authors conclude:

“Economic theory and economic history both tell us that immigration reduced real wages in the host country and that emigration raised real wages in the origin country during the first global century.”[107]

Prior to World War 1, wage impacts were often dramatic: based on CGE simulations of the period between 1870 and 1910, Hatton and Williamson (2006) suggest that the US real wage would have been 34 percent higher than it actually was in 1910 and the British real wage would have been about 12 percent lower had no US immigration or British emigration occurred.[108] These large effects were observed even though there were large flows of capital alongside labour flows, reducing the impact of immigration on the capital to labour ratio in both economies.[109]

4.1.2  Contemporary labour market impacts

Factor price equalisation is always a relevant driver of dynamic markets. However, over the last century the geographic pattern of differentials has changed, and restrictions have been introduced, reducing the impact of equalisation. Further, host countries today have larger populations and per capita incomes, reducing the proportionate impact of any given number of immigrants.[110] Good transport and communications technology and high levels of education allow modern markets to respond quickly. Some economies have flexible labour markets, which also support adjustment to immigration. The new production frontiers of the modern economy are typically in services, where necessary capital such as computers and office space are more easily put in place than in the manufacturing plant of the past.

Contemporary literature focuses more on the impacts of immigration on host country workers. The theoretical predictions depend on the underlying framework. For example, in the simplest standard neoclassical model, where immigrants and locals are perfect substitutes, an increase in immigration will lower the wage paid to local workers. Local labour force participation and employment rates will decrease, and unemployment rates will rise. In other models, including some general equilibrium models, the physical capital stock adjusts in response to immigration, and immigrants are assumed to be imperfect substitutes for (or complements to) locals. The more fully and quickly capital adjusts to changes in labour and the greater the complementarities between local and immigrant workers the smaller and shorter the negative impact of immigration on wages.[111] In further contrast, some contemporary models suggest that material positive labour market impacts can occur through diversity in productive skills and increased innovation.[112] In this framework, immigration can raise the wages of the locals, though in the model effects are economy-wide, so distributional and compositional effects are unseen.

The results of a synthesis of recent empirical work in New Zealand are consistent with international evidence suggesting that immigrants do not have a significant negative impact on the labour market outcomes of the local population.[113] The consensus is that there are only small effects of immigration on the wages and employment of locals.[114] An earlier meta-analysis of 45 primary studies on migration, including some from New Zealand, found that “the impact of immigration on the labour market of the native born is quantitatively very small and estimated coefficients are more than half the time statistically insignificant.”[115]

The underlying reasons for this overall benign impact have been debated for some time and are still being discussed.[116] The consensus is that they likely include short run and long run changes to the output mix, capital levels, domestic labour supply, and productivity which may offset the negative impact of a positive labour supply shock.[117] In other words, the features of the immigrants and the environment they arrived in have been such that adjustment has occurred at relatively low cost.

In the New Zealand context, Hodgson and Poot (2010) suggest that increased local demand from immigrants leads to an inflow of capital which would offset any downward wage pressure.[118] Stillman and Maré (2010) find little evidence that migrant inflows displace either the New Zealand-born or earlier migrants with similar skills in the areas that new migrants are settling, which suggests it is unlikely that internal mobility moderates any potential impacts of immigration on the labour market. The study did not test whether New Zealand-born and earlier migrants were displaced out of the country. In New Zealand, large scale immigration has occurred alongside large scale emigration, which, other things equal, would reduce wage pressure.[119]

Although the overall impact may be small, immigration can still have effects on particular workers and markets, depending on the extent to which immigrants are substitutes or complements to local workers. For example, the meta-analysis conducted by Longhi et. al. (2008) finds a strong and “statistically significant downward effect of newcomers on the wages of earlier migrants, suggesting that in many cases, the substitution elasticity between new arrivals and earlier immigrants will be relatively high.”[120]

Dustmann et. al. (2007) show that in a simple model, a skill group whose relative supply has decreased as a consequence of migration will benefit, while a skill group whose relative supply increases will lose out.[121] The degree of substitutability or complementarity drives the size of these effects. This is confirmed by United Kingdom studies finding increased casualisation of low-wage work, a clear “migrant division of labour” in London, and downward wage pressure for low-skilled workers following the rapid increase in unskilled labour flows associated with the increase in migration from Eastern Europe to the United Kingdom following EU enlargement in 2004.[122]

In the New Zealand context, Maré and Stillman (2009) found that increases in the relative skill composition of migrant inflows had a small negative effect on the wages of high-skilled New Zealand workers. McLeod and Maré (2013a and 2013b) do not find any evidence that the large rise in temporary migration over the past decade has affected monthly wages or total employment in New Zealand, but acknowledge that during most of their sample the business cycle was favourable.

Overall for New Zealand, it can be argued that labour market institutions have responded well to increased immigration since the 1990s, particularly in the light of a starting position of high unemployment and few specific labour shortages. This experience is consistent with theory, and the literature on outcomes elsewhere.

Notes

  • [106]Hatton and Williamson (2006), p. 3.
  • [107]Ibid, p. 7.
  • [108]Ibid, p. 5.
  • [109]Ibid, p. 6.
  • [110]Di Giovanni et.al. (2013) note this influence in their contemporary CGE model, which finds welfare gains across countries from immigration of a little over 2 percent (p. 24), with much larger gains to the migrants themselves (for example, 25 percent for migrants from New Zealand to Australia, p.42).
  • [111]See, for example, Ottaviano and Peri (2006).
  • [112]Ortega and Peri (2013).
  • [113]Hodgson and Poot (2010) looked at studies by Maré and Stillman (2009), Maré et. al. (2007), and Stillman and Maré (2007).
  • [114]There are some studies that demonstrate larger effects, but these tend to rely on complex identification strategies (see Ottaviano and Peri (2008, 2006, 2005)).
  • [115]Longhi et. al. (2008), p. 24.
  • [116]For early examples, see Altonji and Card (1991), Card (1991), Chiswick et. al. (1992), Hanson and Slaughter (1999). Several authors raise the possibility that inappropriate econometric methodology (largely failing to factor in endogenous responses following an immigration shock into models) or labour market rigidities preventing adjustment could lead to low measured wage effects (see for example, Poot and Cochrane (2005), p. 9).
  • [117]For example, Lewis (2005) and Beaudry et. al. (2006) find evidence that there is weaker adoption of advanced technology, which is complementary to skilled labour, in the presence of larger numbers of unskilled workers. This offsets the wage effects of shifts in the proportion of the unskilled workers.
  • [118]Hodgson and Poot (2010), p. 19.
  • [119]Earlier work by Maré et. al. (2007) found that re-migration was relatively more common among migrants with the highest and lowest qualifications. Using data from the Lithuanian Household Budget Survey and the Irish Census following the eastern enlargement of the European Union in 2004, Elsner (2010) found that a one percentage point increase in the emigration rate increased the real wage of men remaining in Lithuania on average by one percent.
  • [120]Longhi et. al. (2008), p. 24.
  • [121]Dustmann et. al. (2007), p. 14.
  • [122]Cook et. al. (2011); Nathan (2011); Nickell and Saleheen (2009); Wills et. al. (2010); Gordon and Kaplanis (2012).
Page top