Immigration and Economic Freedom: Does Education Matter?

Author:Padilla, Alexandre
  1. Introduction

    This paper builds on Padilla and Cachanosky (2018) and examines whether immigrants' educational attainments matter when testing immigrants' institutional impact. Borjas (2014, 2015, 2016) argues that much of the literature examining the global economic benefits of immigration ignores immigrants' institutional impact. (1) His criticism is twofold. First, Borjas (2015, p. 169) argues that this literature ignores the possibility that immigrants coming from countries with low-quality institutions might import their institutions into their host countries, which in turn would deteriorate host countries' institutions. Second, Borjas (2016, pp. 66-87) revives Benjamin Franklin's ([1753] 1904, pp. 408-16) concerns about immigrants' educational attainments and their related inability to assimilate and to understand the importance of preserving the host country's unique institutions. As a result, Borjas argues that a massive migration to the United States of a poorly educated population coming from countries with poor institutions could depress economic growth.

    Most of the attention has been given to the first part of Borjas's criticism of the literature examining the global economic benefits of immigration. This strand of the literature tests what impact, if any, immigrants have on the institutions of their host countries. Using economic freedom as a proxy for institutional quality, (2) most of these studies find statistically and economically significant positive correlations between immigration and the level of economic freedom of their host countries or states (Clark et al. 2015; Powell, Clark, and Nowrasteh 2017; Nowrasteh, Forrester, and Blondin 2019). (3) When they find a negative relationship between immigration and economic freedom, that relationship is neither statistically nor economically significant (Padilla and Cachanosky 2018; Padilla and Cachanosky 2019). However, economists have given much less attention to the second part of Borjas's criticism, that is, whether immigrants' educational attainment matters when it comes to impacting their destination countries' institutions. Even when the aforementioned literature accounts for the quality of the institutions of immigrants' origin countries, it essentially treats all immigrants from these countries as one homogeneous group (Padilla and Cachanosky 2019). In other words, it does not take into account the educational attainments of immigrants. If Borjas (and Franklin) are correct in their conjecture, the "non-effect" that immigrants have on institutions might be the results of two opposite effects neutralizing each other. Therefore, not finding any institutional impact from immigration does not really address Borjas's concerns about the impact unskilled immigrants may have on their host countries.

    Building on Padilla and Cachanosky (2018), this paper fills the gap in the literature. We test Borjas's conjecture and examine whether immigrants with different levels of educational attainment have different impacts on the US states' economic freedom. If Borjas (and Franklin) are correct, we should anticipate a positive relationship between immigrants with higher levels of education and economic freedom. Additionally, we should expect the relationship between immigrants with lower levels of education and economic freedom to be negative.

    Section 2 discusses the data and our model. Section 3 discusses our results. Section 4 concludes.

  2. Data and Model

    Since this paper builds on Padilla and Cachanosky (2018), we follow a similar empirical strategy to assess the institutional impact that immigrants and, more particularly, immigrants grouped according to their educational attainment, have on the US states' economic institutions. We use the levels of economic freedom from the Economic Freedom of North America (EFNA) 2015 report (Stansel, Torra, and McMahon 2016) as our dependent variable. (6) The EFNA report measures "the extent to which policies of individual provinces and states are supportive of economic freedom--the ability of individuals to act in the economic sphere free of undue restrictions" (Stansel, Torra, and McMahon 2016, p. v) The report uses ten variables across three areas: (1) government spending; (2) taxes; and (3) labor market freedom. The EFNA assigns economic freedom scores to sub-areas in these three areas. Among these sub-areas, we are interested in (1) government transfers and subsidies as a percentage of income (Area 1B); (2) top marginal income tax rate and the income threshold at which it applies (Area 2B); and (3) minimum wage legislation (Area 3i). These three sub-areas are often seen as areas that immigration, particularly unskilled immigration, would affect. Since immigrants tend to be less educated and skilled and, consequently, earn less, they might push for policies to increase redistribution toward them.

    The EFNA uses a scale from zero to ten for each component, where ten represents the highest level of economic freedom. In addition, the authors argue that to avoid "imposing subjective judgments about the relative importance of the components, each area was equally weighted and each component within each area was equally weighted" (Stansel, Torra, and McMahon 2016, p. 9).

    Since one of Borjas's concerns is that most new immigrants are less educated and skilled than those in earlier immigration waves were, we group immigrants in three groups: immigrants who did not earn a high school diploma, immigrants with a high school diploma or equivalent, and immigrants who went to college. This latter variable represents all immigrants who went to college. This group includes immigrants having some college but less than one year, immigrants having one or more years of college but no degree, and immigrants whose highest educational attainment is an associate's degree, a bachelor's degree, a master's degree, a professional degree beyond a bachelor's degree, or a doctoral degree. Since the literature shows that better-educated people tend to think more like economists, we would expect to find a positive relationship between immigrants with a college education and economic freedom (Caplan 2001). At the other extreme, we would expect to find a negative relationship between economic freedom and immigrants who did not attend or did not finish high school.

    Our data on immigrants come from the US Census public-use microdata available from IPUMS (Ruggles et al. 2015). We use data from 1980, 1990, and 2000 (5 percent samples) and from the 2010 American Community Survey. Using those data, we construct a panel with data at ten-year intervals, where the starting date of the panel refers to the dependent variable (i.e., t = 2010, so t - 1 = 2000).

    To study the relationship between immigration and economic freedom, we use an empirical strategy similar to the one adopted by Acemoglu et al. (2008) and Spilimbergo (2009): dynamic panel regressions. Our basic model is

    [mathematical expression not reproducible] (1)

    where [ef.sub.it] is the level of economic freedom of a state i in period t. We include the lagged value of economic freedom in our specification in order to capture the various long-run historical, cultural, economic, political, and other factors that influence economic freedom. Our main variable of interest, [imm.sub.iet-1], is the lagged value of the share of immigrants of education level e in state i's total population. Therefore, parameter [gamma] measures the effect of foreign-born immigrants of a specific education level on economic freedom. All other potential covariates are included in the vector [X.sub.it-1]. Finally, [[delta].sub.i] denotes a full set of state dummies and [[mu].sub.t] denotes a full set of year dummies to capture common shocks (common trends) in the economic freedom score of all states; % is an error term, which captures omitted factors, with E([[epsilon].sub.it]) = 0 for all i and t. (7)

    The set of additional controls included in the vector [X.sub.it-1] and likely to affect economic freedom are as follows:

    (i) the share of US natives with a college education

    (ii) the log of personal income per capita (excluding government transfer payments) adjusted for inflation in 2010 dollars;

    (iii) the share of state population living in urban areas; and

    (iv) the share of state population identifying as African-American

    We follow Higgins et al. (2006) and use the personal income measure from the Regional Economic Information System of the Bureau of Economic Analysis (BEA-REIS), which, along with census mid-year state population estimates, gives us per capita income. We adjust per capita income by subtracting government transfers and converting it to 2010 dollars. In all our regressions, we use the log of real income per capita throughout and lag it.

    The share of US natives with a college education comes from the US Census Bureau's public-use microdata as available from IPUMS (Ruggles et al. 2015). That variable contains the share of US natives who have had some college but less than one year, those who have had one or more years of college credit but no degree, and the US natives whose highest educational attainment is an associate's degree, a bachelor's degree, a master's degree, a professional degree beyond a bachelor's degree, or a doctoral degree. As we mentioned earlier, the literature shows that people with a higher level of education tend to think more like economists. Therefore, we would expect to find a positive relationship between immigrants with a college education and economic freedom (Caplan 2001).

    The share of state population living in urban areas and the share of state population who identify as African-American are also from the US Census Bureau's public-use microdata as available from IPUMS (Ruggles et al. 2015). The measure of urbanization is used "to control for the degree of 'cosmopolitanism' in states that may, for various reasons, be more accepting of immigrant inclusion" (Hero and Preuhs...

To continue reading

FREE SIGN UP