Federal Reserve Economic Data

The FRED® Blog

The economics behind the motivation to migrate

Income gaps and inequality in the U.S. and Central America's Northern Triangle

In the past two years, the surge in undocumented immigrants from Central America’s Northern Triangle has been covered extensively by most news outlets. The stories of these migrants from El Salvador, Guatemala, and Honduras involve compelling and often perilous human experiences and intense reactions to the issues involved.

Apart from the political and social views about immigration, there are fundamental questions to ask that may have some economic answers: What is the main motivation for these migrations? And why are people willing to put themselves and their families at great risk to migrate to the U.S.?*

The graph above shows the ratio of per capita income in the U.S., the intended destination for many of these migrants, to per capita income in the three Northern Triangle countries: El Salvador (in blue), Guatemala (in red), and Honduras (in green). The gaps are huge, as expected, but also quite varied, with clear movements over time.

Some history: For El Salvador, during most of the 1970s, the ratio was below 10. But, as a result of the civil war (1979-92), the gap surged to 17.5 by the late 1990s, which coincides with the migration of many Salvadorans to the U.S. (especially to L.A., D.C., N.Y., and Houston.) Ever since, the ratio has remained at the high end of its trajectory, around 15. Guatemala has a similar pattern: The ratio steadily rose from around 11 in 1980 to more than 18 in 2005, and it also has remained at a higher level. In Honduras, the poorest country in Central America, we see even more dramatic disparity: The ratio for Honduras has never been lower than 17.5. It reaches its peak of 28.4 in 1999, and as of today it’s at 25.

The income gaps between the U.S. and the three source countries reveal the magnitude of the potential earnings migrants could gain and the potential improvements migrants could experience in their living conditions. That is, the data suggest that increased migration is motivated by economic considerations. Obviously, these migrants wouldn’t expect, if they managed to enter and remain in the U.S., that they’d attain the average income of U.S. residents. Undocumented workers with much lower labor market qualifications would receive much less than the average. So the ratios in the above graph seem to greatly overestimate income gains. But consider that the countries in the Northern Triangle have traditionally had enormous internal economic disparity, and many immigrants are from the poorer segments of the population. So the ratios could greatly underestimate the earnings gains.

The second graph conveys income disparity by showing the Gini coefficients for El Salvador, Guatemala, and Honduras, as well as for the U.S. (The previous FRED Blog post also used Gini coefficients, a very common indicator of inequality: The higher the Gini, the more concentrated the income distribution: A value of 100% indicates perfect inequality, in the sense that all income would be concentrated with one person [or the tiniest fraction of the population]. A value of 0% indicates perfect equality, a state in which everyone has the same income.)

For most of these years, the Gini coefficients for these countries are very high. Guatemala and Honduras maintain similar levels over time, above 50%, with very slow improvement. In the early 1980s, El Salvador was on par with them; but since the end of the war, its inequality seems to have trended down dramatically. In fact, by the end of the sample, El Salvador exhibits less inequality than the U.S. But it should not be surprising that very poor countries have many desperate people and generate the economic motivation to migrate.

* “I cannot help feeling self-conscious as I try to answer these questions from the comforts of my office. But my aim in this FRED Blog post, as in every other FRED Blog post, is to show how using data from FRED can provide some objective, big-picture perspectives, even on this highly charged issue.” —Alexander Monge-Naranjo

How these graphs were created: For the first graph, search for and select “GDP per capita for the United States in constant dollars” (series ID NYGDPPCAPKDUSA). From the “Edit Graph” panel’s “Edit Lines” tab, use the “Customize data” tool to search for and add “GDP per capita for El Salvador in constant dollars” (series ID NYGDPPCAPKDSLV). Then add the formula a/b. Repeat these steps for Guatemala and Honduras. For the second graph, search for and select “Gini index for El Salvador” (series ID SIPOVGINISLV), and do the same for Honduras, Guatemala, and the U.S. From the “Edit Graph” panel’s “Format” tab, choose “Mark type” square with a width of 5 and a “Line style” width of 1 for all.

Suggested by Alexander Monge-Naranjo.

View on FRED, series used in this post: NYGDPPCAPKDGTM, NYGDPPCAPKDHND, NYGDPPCAPKDSLV, NYGDPPCAPKDUSA, SIPOVGINIGTM, SIPOVGINIHND, SIPOVGINISLV, SIPOVGINIUSA

Chile’s been hot, politically and economically

Comparing growth and equality in South America

If you’ve followed the evolution of Latin America for the past 25 or 30 years, you might be shocked by the recent unrest in Chile: The country’s economic growth has been stunning and widely seen as a harbinger of growth and development for all of Latin America, a region that has underperformed for many years. Chile’s growth is also touted as a confirmation that the openness and liberalization advocated by the “Washington Consensus” was the right prescription for the country and, by extension, for all of Latin America. Of course, this view has had dissenters. But regardless of ideological alliances or inclinations, most were surprised by the depth, pervasiveness, and persistence of the recent violent protests and demonstrations in Chile.

We hope FRED can add something to this discussion, while knowing full well that the issues behind the violence in Chile exist at a much deeper level than the economic data can reach. But the data can provide some interesting context and a starting point.

The graph above shows per capita income in Chile from 1960 until 2018 (thick black line). It also includes the equivalent series for four other major Latin American countries: Argentina, Brazil, Mexico, and Peru (thinner, lighter lines).

It’s easy to see why Chile has been widely considered a success story. Measured in constant 2010 U.S. dollars, Chile’s growth in the 1960s was way below Argentina’s, at par with Mexico’s and Brazil’s, and not far above Peru’s. The 1970s were tough, both politically and economically. By 1975, Chile had fallen way below Mexico and Brazil, and Peru had caught up with it. Another major recession in the late 1970s and early 1980s sent Chile’s income back to its level in 1970. But then Chile started to move away from the pack… By consistently outgrowing its neighbors, Chile caught up with Argentina, Brazil, and Mexico in the early 1990s, leaving Peru way behind. By 2018, Chile’s growth had risen to 50% above Argentina’s and Mexico’s, 40% above Brazil’s, and a whopping 150% above Peru’s.

So, given this apparent economic success, why are Chileans protesting so aggressively? One possibility is that the economic growth has been very unequally distributed.* The second graph sheds some light on this perspective with data from the World Bank. The scale on the left vertical axis shows the Gini coefficient, a very common indicator of inequality, for Chile and its four neighbors. The higher the Gini, the more concentrated the income distribution: A value of 100% indicates perfect inequality, in the sense that all income would be concentrated with one person (or the tiniest fraction of the population). A value of 0% indicates perfect equality, a state in which everyone has the same income. 

A number of interesting patterns emerge. First, the trend for all the countries is a declining Gini; i.e., movement toward a more equal distribution of income. This pattern contradicts the narrative that the Chilean model could deliver growth only by expanding inequality. However, only Brazil, a country widely singled out for its remarkable inequality, exhibits a Gini coefficient substantially higher than Chile’s. Otherwise, this rough measure of inequality puts Chile on par with Mexico (another country singled out for its inequality). Interestingly, both Mexico and Brazil have had massive social unrest in the past two years. Finally, the Gini coefficients indicate that inequality in Chile is substantially higher than in Argentina and, surprisingly, Peru.

Obviously, these simple patterns cannot fully explain the recent protests. If they could, no one would have been surprised by them. However, the data can help add some context and a bit of discipline when considering and challenging popular narratives.

* “The typical narrative inside the country when I was living there in the early 90s, and the narrative in Latin America in general, is that income inequality is very high in Chile.” —Alexander Monge-Naranjo

How to make these graphs: For the first graph, search for and select “GDP per capita for Chile in constant dollars” (series ID NYGDPPCAPKDCHL); click “Add to Graph.” From the “Edit Graph” panel, use the “Add Line” tab and add similar series for Argentina, Brazil, Mexico, and Peru. From the “Format” tab, choose line thickness “5” and color “black” for the Chile series. For the second graph, search for and select the Gini series instead of the GDP per capita series. From the “Format” tab, select “Dot” under “Line style:” because data are not available for every year.

Suggested by Alexander Monge-Naranjo.

View on FRED, series used in this post: NYGDPPCAPKDARG, NYGDPPCAPKDBRA, NYGDPPCAPKDCHL, NYGDPPCAPKDMEX, NYGDPPCAPKDPER, SIPOVGINIARG, SIPOVGINIBRA, SIPOVGINICHL, SIPOVGINIMEX, SIPOVGINIPER

The rich borrow, too

Liability distribution across rich and poor households

Over the past  few  weeks, we’ve used data from a dataset compiled by the Federal Reserve Board specifically to analyze the distribution of data across households. While our target so far has been assets, today we look at liabilities. How do rich and poor households borrow?

The graph above shows the total liabilities of four wealth classes: the top 1%, the next 9%, the next 40%, and the bottom 50%. At first glance, it appears that richer households hold less in liabilities. But if you hover over the graph, you see the actual percentages: the top 1% hold 4.6% of all liabilities, the bottom half 36%. One might assume the rich borrow less and the poor borrow more. But to better understand this, let’s take a look at the major categories on the liability side of things.

The first category is mortgages, shown in the second graph. Poorer households are less likely to own a home, and when they do it is a smaller home. As they have less funds, they need proportionally larger mortgages to own those smaller homes. Richer households need smaller mortgages, but they have a fiscal incentive for larger mortgages, as mortgage interest is deductible from their taxes and their tax rates are likely higher. In the end, we see that the top 1% hold 4.2% of all mortgages, while the bottom half has 39%.

The second category is consumer credit, on shown in the bottom graph. This includes credit cards, student loans, car loans, and other similar liabilities. Here the bottom half amasses 54% of the total debt; quite obviously they borrow to make many purchases. But the top 1% also punch above their weight with 2.1% of total consumer credit. How come? One reason is that expensive but highly rewarding courses of study (medicine, law) contribute quite a bit to student debt. Also, the consumer credit numbers include credit card balances paid in full (about 30% of credit card debt, or 7% of all consumer credit).

How these graphs were created: The procedure is the same for each graph. You can find these series in the Distributional Financial Accounts or the Levels of Wealth by Wealth Percentile Groups release table; check the series you want, and click “Add to Graph.” From the “Edit Graph” menu, open the “Format” tab to choose graph type “Area” with stacking “Percent.”

Suggested by Christian Zimmermann.

View on FRED, series used in this post: WFRBLB50101, WFRBLB50102, WFRBLB50103, WFRBLN09047, WFRBLN09048, WFRBLN09049, WFRBLN40074, WFRBLN40075, WFRBLN40076, WFRBLT01020, WFRBLT01021, WFRBLT01022


Subscribe to the FRED newsletter


Follow us

Back to Top