The Effect of Employment Opportunities on Migration Trends Across Countries

By: Ashley Holst

http://chartsbin.com

Migration is a natural part of our globalized world. People migrate for an endless number of reasons. Some migrate for short periods only to return to their home countries, while others migrate permanently in hopes that they may better their economic status and improve the livelihoods of their families at home through remittances. Improved financial status and the quest for a larger or more stable income is one of the most common pushes for migration.  Labor migration is defined as the “movement of persons from one state to another… for the purpose of employment” (IOM[1]). An economic migrant may have more freedom of movement than a forced migrant or a migrant seeking refuge from persecution, but economic migrants also often feel forced to migrate in order to escape poverty and unemployment. In developing and middle developing countries unemployment, lack of opportunities, and high levels of poverty lead to increased emigration (ILO[2]). In a briefing on labor migration, the OECD[3] states “Regarding the labor market, migration is a symptom of imbalances in sending countries, such as high rates of unemployment and underemployment among low-skilled workers, low wages for skilled workers, and unmet demand for education and acquisition of skills” (2009). Despite the widespread agreement that lack of economic security and opportunities for employment lead to emigration, there is little qualitative data on the effect of lack of employment opportunities on migration.

What is the effect of lack of employment opportunities on migration patterns?

The primary dependent variable for this model is Net Migration[4], while a clear measurement of outmigration would have been ideal, net migration offers a sense of both immigration and emigration in a country. The primary independent variable is labor force participation[5]. This variable measures the percentage of the population ages 15-64 that are currently employed. By using this variable, I assume that if employment opportunities are available that people will take advantage of them.

To measure the effect of labor force participation, I wanted to look at a wide variety of countries. It would have been incomplete to answer this question only for developing or developed countries and I believe that migration trends will follow employment trends across countries, not only within countries. As examples, I observed trends in Haiti, one of the poorest countries in the world and India, one of the world’s rising economic powers. These two countries also lend themselves to a comparison of employment and migration, as Haiti is known for its poverty and unemployment issues while India is a growing hub for companies. 

As seen in the graphs, migration trends seem to follow labor force participation rates. In Haiti, labor force participation rates fell between 1990 and 2000 resulting in increased outmigration in the following years, when labor force participation began to increase, outmigration was quickly reduced. The 2010 earthquake may have influenced the spike in net migration, but the trends prior to 2010 still reflect the labor force participation rates of several years previous. The observations in India show similar results, prior to 2000 labor force participation was decreasing and net migration was decreasing. After a reverse in direction of labor force participation, migration too changed directions showing a higher net migration.

Three additional variables were included in the model. Each of the control variables is a possible factor influencing migration trends: Level of Democracy, presence of conflict and GDP per capita[6]. The summary statistics for each of the variables included in the model can be seen below. I expect that level of democracy, conflict, and GDP per capita have effects on both labor force participation and net migration, thus I control for them in the regression analysis. The years included in the model are 1990 to 2012.

I use cross sectional time series panel data and a between estimator to measure the strength of the effect of labor force participation on net migration. A between estimator uses an Ordinary Least Squares regression on the mean values of the independent variables on the mean value of the dependent variable, net migration, over a given time period. This regression is useful in cases where variables are slow changing or “sluggish.” Doing a between estimator over time also reduces measurement error. Due to the lack of measurements available, between effects offers the best analysis of the strength of the effect of labor force participation on net migration.

The results in column one show that Labor Force Participation does not have a significant effect on net migration but the relationship does show the expected direction (increased labor participation resulting in a decrease in net migration which indicates increasing outmigration). The second and third columns show that GDP per capita, level of democracy, and conflict are also not significantly correlated to net migration. Increased democracy results in higher net migration (less people leaving) while an increase in the severity of conflict resulted in lower net migration (more people leaving).  GDP per capita and net migration have a non-linear relationship where above a certain level of GDP per capita, net migration begins decreasing. This could be due to the fact that it is more difficult to immigrate to countries with higher GDP, thus affecting the net migration score.

The fourth column controlling for all the variables, showed significant results for GDP per capita and conflict.  When controlling for the other variables, the direction of the relationship for GDP and GDP-squared became the reverse while conflict maintained direction. The primary independent variable, Labor Force Participation rate remained negative and not statistically significant.

The results show that migration is fluid and that there are many factors that have an effect on net migration rates. People chose to stay, leave, and where to migrate to based on economic, social, and physical factors. Labor force participation, or the availability of sufficient employment opportunities does not significantly affect migration patterns. This result could be due to the fact that people not participating in the formal labor markets may be earning incomes in other informal ways. It is also possible that the exclusion of undocumented migration may be playing an affect in this regression. It would be interesting to do a similar model looking at documented versus undocumented migration and these same influencing variables; the results may differ between the groups. Unfortunately accurate data on undocumented migration is difficult to obtain and would likely include significant errors. In light of these results, countries suffering from high levels of emigration should focus on reducing conflict within their country and improving GDP per capita levels.

---

[1] http://www.iom.int/key-migration-terms

[2] http://www.ilo.org/wcmsp5/groups/public/---ed_protect/---protrav/---migrant/documents/publication/wcms_178672.pdf

[3] http://www.oecd.org/development/povertyreduction/43280513.pdf

[4] Net Migration: Measured as the number of entering immigrants minus emigrants. Positive numbers reflect more immigrants than emigrants, while negative values reflect more emigrants leaving than immigrants entering the country. Net migration is logged to correct for positive skew.

All data collected from the Quality of Governance Standard Cross-Sectional Time Series Data Set (2015) http://qog.pol.gu.se/data/datadownloads/qogstandarddata 

[5] Labor Force Participation Rate: the proportion of the population ages 15-64 that is economically active: all people who supply labor for the production of goods and services during a specified period.

[6] Freedom House Polity measures level of democracy on a 0 to 10 scale. Internationalized internal armed conflict measures conflict within a country between governments or opposition to government that requires outside intervention. Conflict is scored on a 1 to 10 scale where 1 indicates no violent incidents and 10 indicates civil war or widespread conflict.  GDP per capita PPP is in $US and based on a 2005 based year. GDP per capita is logged and squared to control for skew.