Labour market equilibrium
Does migration change the equilibrium of the economy, and more specifically, does it have an impact on the labour market? The analysis of the macro effects of emigration can be seen in two ways: by looking at transmissions between households and by looking at aggregate macroeconomic indicators. Microeconomic models focusing on individual actors (individuals, households or firms) miss the linkages that transmit influences among households. The effects on the economic agents who are directly involved in migration are only part of the story of how migration reshapes rural economies. Direct remittance effects generally are small compared with indirect ones.
Evidence is provided by a simulation exercise by Dyer and Taylor (2009) on rural Mexican households, showing that emigration raises wages in communities where transactions are high between households. Specifically, their analysis shows that emigration tends to increase wages. Changes in labour availability and the increase in remittances impact wages and the prices of non-tradeables, and consequently generate general equilibrium effects on production, incomes and investments by households that have not sent migrants abroad. They also find that the landless and smallholder households benefit from the higher wages that are induced by the new investments in production activities by migrant households.
According to a general neoclassical labour and demand approach, a decrease in workers should lead to an increase in wages. There are two reasons for this. First, if a significant share of the economically active population leaves the country, the immediate drop in labour supply should lead to an increase in wages, at least in the short term before capital adjustment occurs. But remittances, as has been seen, also reinforce the impact if they lead to migrant household members working less, further decreasing the supply of labour - but without people actually leaving the country.
Complicating the issue slightly is the fact that workers with similar characteristics will affect each other's wages and employment chances and will follow similar market responses. The effect of complementary workers will follow in the opposite direction. In other words, there are many labour markets depending on individual skill and experience levels. For example, if many agricultural workers arrive at once, it will apply downward pressure on agricultural workers in that region, but as production increases, so will the demand for truck drivers bringing produce to the market.
The issue is still a recent and contested one for empirical economists. The link between emigration and the equilibrium of the labour market in developing countries has a few added complexities in comparison to immigration and to developing countries:
- • First, as seen earlier, a number of developing countries receive large amounts of remittances, sometimes representing more than 20% of GDP (for instance Lebanon, Lesotho, Moldova and Tajikistan).
- • Second, upward pressure may mount on wages and spur j ob creation. But, what happens in countries where certain sectors and skill levels have an immense reserve army of unemployed or under-employed?
- • Third, labour market equilibria are highly influenced by the relative amount of capital and labour in the economy. Despite the benefits to labour it brings through the rebalancing of this ratio (labour becomes more scarce), the country loses since it scales down the overall labour component of its production function. That is, the overall productive capacity of the country diminishes.
The anecdotal evidence provided by Macharia (2003), on Kenya, and Ennaji and Sadiqi (2004), on Morocco, suggest that wages do increase as a result of emigration.3 Can this be tested empirically? Several approaches can be taken to answer this question. One is to look at whether emigration contributes to wage convergence in richer countries in the long run. In historic terms, it has been argued that emigration was a contributing driver for real wage convergence to that of richer countries by helping decrease the growth of the labour force in the home country. Such was the experience, for instance, of many European countries before World War I (Boyer et al., 1994; Williamson, 1996). This approach, however, requires a very long-term view. Complicating the issue is that in some countries, emigration may have also altered the structural capital-labour ratio.
Most of the recent empirical research takes inspiration from the body of work focusing on the impact of immigration, notably on Borjas (2003) and more recently extending the approach to measure the impact of emigration. In empirical terms, using an approach similar to Borjas (2003), several studies conclude that emigration did indeed increase wages in Mexico (Mishra, 2007; Aydemir and Borjas, 2007), Puerto Rico (Borjas, 2008) and Moldova (Bouton et al., 2009). Using a slightly modified spatial approach to exploit regional differences in Mexico, Hanson (2007b) yields a similar conclusion.
The typical elasticity found in these studies ranges from 2% to 6% (a 10% increase in emigration leads to a 2% to 6% increase in wages). While in Hanson (2007a) it is slightly higher, his methodology does not rule out any indirect effects of emigration (the impact on growth) and therefore probably overvalues the true elasticity. A simulation exercise based on a 1998 social accounting matrix in Morocco also finds that the direction of the effect is positive (Decaluwe and Karam, 2008).
Many of these studies focus on countries that have a relatively wellfunctioning labour market and sustained growth. Moreover, they either look at the very long term or use historical data. What happens in countries where the labour market and economic growth are poor? Do the mechanisms work in the same way or does a reserve army of workers keep taking up the available jobs? What happens when a country is hit by a sudden shock, sending many workers away? In 1998, the track of Hurricane Mitch split Honduras in two and devastated infrastructure along its path. Economic activity ceased and many fled to the US. Box 4.4 provides the results of a study looking at the impact of emigration on Honduran wages.
Box 4.4. Hurricane Mitch and the impact of emigration on wages
Honduras provides an interesting case study for the analysis of the impact of emigration on wages. In 1998, the second deadliest hurricane on record at the time swept through the country and prompted a wave of emigration. The surge in emigration was accompanied by an intense debate on its impact for the country's development, featuring campaign slogans such as "quedate con nosotros" ("stay with us") by the Honduran Association of Maquiladoras. However, while emigration may negatively have affected the maquiladora industry, it benefited Honduran workers staying behind by reducing pressures in the crowded labour market. In fact, average wages increased for workers in Honduras as a result of emigration.
To put the study in context, Honduras is a poor country with a population of fewer than 8 million inhabitants. GDP per capita is low, just over USD 4 000, somewhere in the middle of the ranking in Central America and in 2006, 60% of Honduran households were living under the national poverty line (ISACC, 2009). The labour market is highly segmented between the few existing formal jobs and the rest. Internal migration is mostly undertaken to obtain an opportunity to emigrate internationally and starting a formal business is complicated. Despite the less-than-ideal conditions in the labour market, women have increasingly entered it over the years. Until the hurricane hit, emigration from Honduras was relatively low in comparison to its neighbouring Central American countries; most movement out of the country in the 1970s and 1980s was primarily spurred by regional conflict.
According to the neoclassical model of labour, when there are many individuals with similar jobs leaving the labour market simultaneously, it experiences an upward pressure on wages. To analyse whether emigration has had an impact on wages in Honduras over the years following the hurricane, a Borjas-type of empirical framework is employed using data from bi-annual labour surveys in Honduras and partial censuses in the US for the years 2001, 2004 and 2007. The analysis shows that the sudden wave of emigration from Honduras following Hurricane Mitch yielded an increase in wages of around 10% for every 10% shift of labour supply due to emigration over these years - an elasticity much higher than the studies on other countries referenced above. But the impact diminishes over time - suggesting that capital re-adjusts to the available labour in the country, and the labour market finds its equilibrium in the long term.
The implications are interesting for policy making. As mentioned earlier, the release of pressure in a crowded labour market benefits those staying behind, even though the loss of labour decreases the overall resources available to the country. Moreover, by increasing domestic wages the country takes a step towards converging towards the economic heights of richer countries.
But what categories of workers are driving these results? Intuitively, the categories and education-experience groups with the highest emigration rates should benefit the most. This is true when education is looked at. Of four education groups, the highest emigration rates were by the highest educated followed by the lowest educated. While all education groups gained, these two groups experienced the highest gains in wages.
But it is not always necessarily the case that groups with higher emigration rates gain, for two reasons. First, categories of workers are mobile, in the sense that they can exploit job opportunities left by other categories. The rise in the number of women working in Honduras is a good example. In the study they gained much more than men, because more men were migrating. As many jobs do not necessarily discriminate between men and women, the latter were consequently left with more and better-paying jobs.
Second, the labour market is complex, and certain categories may benefit from more or less opportunity due to their intrinsic character. For instance, in the analysis above, wages in rural areas increased faster than those in urban areas, not because more rural workers left, but because incomplete labour markets imply greater difficulty in replacing workers. Similarly, workers in the private sector saw their wages rise faster than those in the public sector, not because they emigrated in higher numbers but because the (mostly) informal private sector is more reactive to economic changes than the more rigid and protected public sector.
While the emancipation of women is a good sign for development, the results
above reveal potential problems in the Honduran labour market:
- - First, female gains there may be more of a sign that they are taking on too much responsibility. The results hide the fact that while women have taken on a bigger economic role in society, they have not necessarily reduced other burdens.
- - Second, the fact that rural areas are gaining the most means that labour markets there are highly imperfect; agricultural help is costing more because no one is left and farmers are thus losing out.
- - Third, postsecondary educated workers are already those that have the highest returns to labour; an increase in their wages is only increasing inequality between skill groups. Mishra (2007) shows that emigration did indeed increase inequality between education-experience groups in Mexico from 1970 to 2000. Interestingly it is not only the most skilled that gained, however. It is the relative size of groups that matters. Being smaller than those with primary and high school education, the group with no formal education benefited the second most after the university-educated. Interestingly, many of these individuals work in the maquiladora industry, where resistance to emigration is still quite high.
Note: Based on Gagnon (2011).
The analysis cited earlier by Dyer and Taylor (2009) suggested that emigration effects transmit from migrant households to non-migrant households through economic interlinkages. Taking this analysis one step further, migration policies in other countries can impact on migrant-sending countries, and even spill over on to non-migrant households through the same mechanism. These externalities are seldom taken into account.