Economics of Temporal Migrations


Many migrations are temporal, a fact that is often ignored in both the public discourse and the economic literature on migration. Such omission is serious as temporariness of migration affects almost every aspect of immigrants’ life cycle decisions as well as the composition of the remaining migrant population. Failure to include temporal migration into consideration will likely drive to fallacious conclusions due to the dimensions of the phenomenon. A recent OECD report points that, depending on the country and time frame studied, 20 to 50 percent of immigrants leave the host country within the first five years after arrival.[1]

Before proceeding to the topic of temporal migration, let’s step back to distinguish migrants by reason of migration. One can claim that immigration occurs for two reasons: driven by economic motives or by the necessity to escape the country of origin.  Despite increased attention to the second type of migrants, which we shall call refugees, most migrants move from their country of origin for economic motives. Because of this, this analysis will centre in the kind of immigrants who choose to move in search of a job, escape poverty, gain a set of skills they find valuable and so on. Conceptually these two types of migration are very different. Here I shall attempt to elucidate some of the features that make temporal, economic driven migrants so different.

Outmigration in not consequence free. One obvious implication regards the selection of migrants and thus the composition of migrants. That is, migrants who decide to stay are more likely to be similar across them than across those who decide to migrate (on average, at least). Then, one important question is which immigrants return and which migrants stay. To put it differently, are migrants positively or negatively selected? This will depend mainly (while not exclusively) on two main features: before migration skills and re-emigration plans. Of course, plans and ex-post realizations often differ. However, choices made during the stay of migration are largely determined by the expectation of the temporariness of their migration. Moreover, it has been found that expected migration length and actual duration realization is not very different.

Empirical literature has found that intended duration of migration affects assimilation efforts and consumption patterns which in turn affects employment outcomes too. By assimilation efforts one can think of willingness to learn a language, efforts to extend social network, investment in valuable local skills and so on. For example,  Christian Dustmann has shown that those with non-permanent intentions to stay do invest less in language capital, as opposed to permanent migrants.[2] It is not hard to see how assimilation profiles is correlated with employment prospects and thus earnings growth.

Evidence shows that temporal immigrants have flatter earnings profiles than permanent ones.[3] Regarding consumption, economic analysis suggests that temporal migrants are more likely to over-save during migration while permanent migrants spread consumption more equally over their life cycle.[4] Importantly, because temporary immigrants are more likely to leave their families behind, temporary migrants are more likely to hold assets in their home countries and thus to send more remittances too. Such transfers may have important consequences for the home country.
Remittances Banco de MexicoArguably, the most important channel by which migration affects the sending country is via remittances (or transfers from the immigration country to the immigrant’s home country). For example, according to Mexico’s central bank (see figure), reported Mexicans overseas sent nearly $24.8 billion home in 2015, overtaking oil revenues for the first time as a source of foreign income.

Migration can have negative effects for the sending country too. One obvious channel is via the brain drain effect. That is, by a reduction in the per capita human capital in the home country. However, this can be mitigated and even reversed by returning migrants if their gained skills abroad are highly valued in the country of origin. On the other hand, temporary migrants tend to spend their most productive years in the host country, while spending costly childhood and retirement years in the country of origin. This effect is opposite for the receiving country, as the high fiscal burden that aging brings about is borne by the country in which migrants settle for retirement. It is estimated that average annual social expenditure for working-age population, in OECD countries, is twice as high for children and six times as high for individuals over 65.

Why temporal migration occurs introduces additional complexities to the analysis. One can think of circular migrations, as is often the case with Mexican agricultural migrants in the United States, student migrations, or migrants who wish to save for later investments in their home countries. However, in almost every case in which the decision to re-emigrate is made endogenously by the migrant and is free of unexpected shocks, the reason falls back to an intrinsic preference for their home country or to a high return for the skills acquired during migration in the country of origin. Intuitively, the decision revolves around the trade-off a migrant faces between the benefits of staying longer and the loss of foregone utility from spending additional time in the host country.

Implications of temporary versus permanent migrations follow directly from the behavioural consequences between the two types and to selection of out-migration. Temporary migrants may invest less in social and human capital, having potentially large consequences for their assimilation and earnings profile. Temporary migrations are also likely to affect wages and employment of natives in destination countries differently for permanent migrations, how this occurs will depend on who emigrates, and where emigrants are located along the native skill and wage distribution. Of course each case will vary in both the time and space dimension for which any attempt to make universal conclusions needs to be taken with caution. Here I attempt to stress the fact that return migration is perhaps a norm rather than an exception a fact that conveys large consequences for both the sending and receiving countries.


by Diego de la Fuente



[1] Organisation for Economic Cooperation and Development. International Migration Outlook 2008. OECD publishing

[2] Dustmann, C. (1999).  “Temporary Migration, Human Capital, and Language Fluency of Migrants.” Scandinavian Journal of Economics 10.2: 297-314

[3] Dustmann and Gorlach (2015). “The Economics of Temporary Migrations.” Journal of Economic Literature (forthcoming). 

[4] Bauer, T. and Sinning, M. (2011) “The Savings Behaviour of Temporary and Permanent Migrants in Germany.” Journal of Population Economics 24.2: 421-499

Stratification of Poverty in Mexico


In Mexico, 80% of its population is deprived of at least one basic social and economic life dimension. Poverty in Mexico, regardless of how one measures it, is a widespread phenomenon that has grown on a par with population.

The poverty measure employed here focuses on seven categories, one of which is income, four others are cutoff indexes indicating whether individuals have access to health, education, food and social security and two others related to the quality of housing and access to basic housing services (such as running water or electricity). With these categories, a classification scheme for multidimensional poverty is constructed:

  • Extreme Multidimensional Poverty: Income below the necessary income to afford the most basics of life and being deprived of three or more dimensions.
  • Multidimensional Poverty: Income below the necessary income to afford the basics of life and being deprived of one dimension.

Distribution of poverty 2014 - Coneval - 3

Despite this graph showing data of 2014, if one explores previews data it will find that little has changed in the composition of poverty over time. This is highly problematic as only one in five Mexicans live out of poverty hinting to the fact that social mobility is close to nonexistent.

Intergenerational transmission of social and economic conditions is perhaps in the core of understanding Mexico’s economic underperformance. When intergenerational transmission of poverty is elevated, many high skilled individuals will not fully materialize their potential and thus their contribution to society will be suboptimal. Thus, the degree at which individual’s social and economic development level is shaped by non-individual decisions partially determines how resources and opportunities are allocated. This in turn affects long run economic dynamics.

How we understand poverty can importantly influence how we come to measure it, how we analyse it and how we create policies to fight it. Poverty in Mexico is a critical social problem that Mexican governments have widely ignored and failed to tackle.

by Diego de la Fuente


Note: Data on poverty gathered from Coneval 2014.

Poverty in Mexico II


Foto-post-mexicoMexico’s deep roots of poverty which have been systematically ignored by both the political and social spheres throughout its history.  Over 23 million people live with an insufficient income to buy the minimum basket of food, even if they disposed all their income to do so; over 60 million cannot afford the minimum basket of goods and services.

Poverty is unevenly distributed across the population: with the indigenous, the rural and the youth populations being the most severely affected. In terms of the multidimensional poverty measure elaborated by CONEVAL which captures income together with six other indicators (access to food, edu
cation, housing, health and social security and basic services -such as water and sewage-), in the year 2012, poverty affected:

  • 72.2% of the indigenous population (vs. 42.6% of the non-indigenous population).
  • 61.6% of the rural population (vs. 40.6% of the urban population).
  • 53.8% of the under 18 years old (vs. 40.7% of the population between 18 and 65 years).

Poverty is hard to contextualize because more than a concept it is a way of life that is suffered distinctly across populations: in the rural and urban areas, by different age groups, across sexes, regions, climates, cultures, and so on. Some additional numbers should serve to make the point.



    • Illiteracy rates are over four times higher than for the rest of the population, with 27% of the indigenous suffering it.
    • According to UNICEF, 33.2% of under 5 years old presented stunted growth.
    • Mortality rate for indigenous children is over 60% higher than for the general population.


    • According to CONAPO, in the mainly rural state of Chiapas (58% of the population live in towns with 5,000 or less inhabitants): 37% of the population did not finish primary (elementary) school, 70% earn an income below two minimum wages, 25% live in housing with dirt floor, 18% cannot read or write.
    • In the state of Guerrero (49.7% of the population live in towns with 5,000 or less inhabitants): over 20% of the population lived without access to sewage, 29.8% didn’t have access to tubed water, 20% lived in dirt floor housing, 17% cannot read or write.

Children and youth:

The problem of children is as obscure as those of the other groups but in a different way.

    • infantil3According to INEGI, around three million Mexicans between 5 and 17 years old work (12.5% of the total), of which more than two thirds are boys. Of these, 44% do not receive an income, 28% earn an income below the minimum wage and only 8% earn more than 2 minimum wages for their work.
    • Furthermore, over 40% of these children work to complement family income or work; over 25% do so to pay for schooling materials and own expenses and only 5.1% did so to avoid schooling.

In Mexico, life is much harder for some that for others. However, it is striking is to find that poverty in Mexico is, more often than not, the result of exclusion. Looking at the level of need that large groups of the population have points towards the cyclical transmission of the problem across generations.

Paradoxically, the missing component for the country’s long waited economic flourishing lies precisely in whether the country is capable or not of unleashing individual potential across every group of society. Giving opportunities for development to the population would not only create a more just society but a much more dynamic one too.

by Diego de la Fuente

Poverty in Mexico I

In the 1.96 million sq. kilometers of Mexican territory over 120 million people coexist. This is a country of great social, cultural and natural diversity such that is becomes hard to speak of a one Mexico. Along the territory, 67 stable languages are spoken among its various indigenous groups, making it the seventh most linguistically diverse country in the planet. According to CONEVAL (a Mexican government agency that evaluates social policy and measures poverty), around 9.7 percent of the Mexican population speaks an indigenous language. Heterogeneities in the welfare dimension abound too, with poverty being a widespread, yet differently lived, phenomenon.


Last official measures of income (2012) suggested that 51.6 percent of the total population lived with incomes below the poverty line. A level of poverty well above that of any other country with a similar average income. In fact, the share of the total population living under income poverty has not been significantly altered in the last 20 years as it affected a similar 53.1 percent in 1992. This is not to say that poverty has not increased: because of population growth there were more that 20 million additional people living under this type of poverty in 2012.

This is a reflection of the profound levels of opportunity and income inequality of the country. According to the OECD, a club of mostly rich countries, the top 10 percent income earners in Mexico held an income 30.5 times higher than the poorest 10 percent, in contrast with the 9.6 OECD average. Other measures portrait the same problem.

10354745_699662983422490_6674093125968396564_nBut not all groups are affected the same. Not only is poverty deeper in some regions of the country but it is also more widespread. For example, despite the fact that indigenous groups constitute little less than 10 percent of the total population, they accounted for 30.2 percent of the total population living under extreme poverty and 15.4 percent of the population living under general poverty (see graph below: “Poverty Levels Within Municipalities”). More worrisome is the fact that despite this being publicly known, in 2013 only 1.8 percent of the Federal budget was targeted to the development of indigenous towns and communities.

Poverty is the source of much of the current widespread social frustration as its origin has more to do with social exclusion, lack equal opportunities and an economic policy that lies far away from the multifaceted Mexican reality.

New figures of poverty are expected to be released in the coming weeks. Some expect the numbers of poverty to somewhat fall, specially those of extreme poverty since Federal programs such as “Sin Hambre” (Without hunger) -aimed to give basic food supplies to millions in need- have increased under the current administration. However, it is not clear their impact will be large and even if it is, the programs lack the capacity to break poverty in a structural manner. Some rightfully complain they are nothing more than meager palliatives.

Mexico has been used to tackling the problem in a rather superficial way, with doses of hurtful populism that are not rare to see in Latin America. Social policy needs to be better directed to break poverty in an organic manner. It should prioritize on those who need it the most; on those who have been excluded from the development of the economy for a long time. A policy for economic, social and cultural inclusion ought to be a priority.



Additional numbers:

Figura 3 (.wmf) - Indigenismo, desigualdad y pobreza en México

The following graph shows the distribution of municipalities according to the incidence of poverty within municipalities (vertical axis). Municipalities have been divided by the share of the population that speak an indigenous language. Municipalities are considered “indigenous” when over 70 percent of the population speaks an indigenous language. This graph allows us to see that in 75 percent of “indigenous” municipalities at least 80 percent of the population lives under (multidimensional) poverty while the exact reverse is true for the rest of municipalities. Something similar happens with extreme poverty: whereas in 50 percent of the indigenous municipalities over half of its population lives under extreme poverty in the corresponding in half of non-indigenous municipalities less than 15 percent of the population does.

Revolutionary Demographics

Demography has uninterruptedly changed the course of human history. From the diaspora that led the formation of Ancient Greece to the aging populations of Japan and Europe. However, it is possible to assert that the speed of the demographic changes that North America has been going through is unprecedented.

The biggest change in this respect has been migration from the southern countries of Latin America to northern ones, particularly the United States. Not surprisingly, a large proportion of them come from Mexico since it accounts for over 30% of the Spanish speaking population in the Americas and shares a 3,145 km (1,954 mi) border with the United States, in many respects the most dynamic border in the world.

In fact, about one in ten Mexican citizens, 12m in total, live in the United States and despite ugly and divisive yet apparently fashionable political rhetoric half of them are legal and documented migrants. This makes for the biggest immigrant community in the world.

Has this migration come to a reach a new equilibrium? Between 1995 and 2000 some 3m Mexicans moved to the United States, vastly outnumbering the 700,000 or so who returned to Mexico. Yet in 2005-10 the number of newcomers fell to 1.4m, whereas that of returners increased to a matching 1.4m. Some estimates report that current net migration between these two countries has reversed. Much of this is likely to be the result of scarcer job prospects in the US, but not all.

Still, every year around 900,000 United States born citizens with Latin American heritage reach voting age. The effects of past migration are inevitable. The map that follows should help you visualize the size of the phenomenon.


Hispanic Population % 2011 (US Total) Inverted 2



It is not uncommon that people find a connection between the previews map and that of the mid 19th century before the territorial disputes between the United States and Mexico that were settled between 1845 and 1848 to favor the rising US power. The map to the right plots the Mexican territory as claimed by Mexico.


by Diego de la Fuente