The main objective of this research is to estimate internal and external wage effects associated with a changing share of college graduates in Canada. This paper uses data drawn from the Canadian 1991, 1996, and 2001 Public Use Microdata File for 25 to 65 years old individuals working full-time and full-year in metropolitan areas. These workers are then separated into four different levels of education groups in order to estimate the effect of change in the share of college-educated workers on their earnings. <p>The Ordinary Least Squares (OLS) estimates, controlling for potential work experience, total years of schooling, individual occupation, employment industry, immigration status, visible minority status, show a significant positive relationship between the percentage change of the share of college-educated workers and the percentage change of individuals real weekly wage rates. We found that one percentage point increase in a census metropolitan areas share of college-graduated workers was associated with a 0.35 percentage change in all workers wage rates in that city. For separated education groups, our results showed that a one percentage expansion in the supply of college-graduated workers raised less than high schools wage rate by 0.245 percent, raised high-school graduates wage rate by 0.363 percent, raised more than college-educated workers wage rate by 0.385 percent, and raised college-educated wage rate by 0.326 percentage. These results are consistent with the conclusion arrived at by E. Moretti, (2004) that all types of workers earnings increased when a citys share of college graduates rose.

The main objective of this research is to estimate internal and external wage effects associated with a changing share of college graduates in Canada. This paper uses data drawn from the Canadian 1991, 1996, and 2001 Public Use Microdata File for 25 to 65 years old individuals working full-time and full-year in metropolitan areas. These workers are then separated into four different levels of education groups in order to estimate the effect of change in the share of college-educated workers on their earnings. <p>The Ordinary Least Squares (OLS) estimates, controlling for potential work experience, total years of schooling, individual occupation, employment industry, immigration status, visible minority status, show a significant positive relationship between the percentage change of the share of college-educated workers and the percentage change of individuals real weekly wage rates. We found that one percentage point increase in a census metropolitan areas share of college-graduated workers was associated with a 0.35 percentage change in all workers wage rates in that city. For separated education groups, our results showed that a one percentage expansion in the supply of college-graduated workers raised less than high schools wage rate by 0.245 percent, raised high-school graduates wage rate by 0.363 percent, raised more than college-educated workers wage rate by 0.385 percent, and raised college-educated wage rate by 0.326 percentage. These results are consistent with the conclusion arrived at by E. Moretti, (2004) that all types of workers earnings increased when a citys share of college graduates rose.

Submitted by Francisco Augusto Seixas Thomé (francisco.thome@fgv.br) on 2013-01-03T21:38:35Z No. of bitstreams: 1 Thomé.pdf: 1674130 bytes, checksum: b3ec4566d7dfee1d5ea069cfedaa082a (MD5) / Approved for entry into archive by Vitor Souza (vitor.souza@fgv.br) on 2013-01-15T13:09:01Z (GMT) No. of bitstreams: 1 Thomé.pdf: 1674130 bytes, checksum: b3ec4566d7dfee1d5ea069cfedaa082a (MD5) / Made available in DSpace on 2013-02-04T18:21:25Z (GMT). No. of bitstreams: 1 Thomé.pdf: 1674130 bytes, checksum: b3ec4566d7dfee1d5ea069cfedaa082a (MD5) Previous issue date: 2012-05-31 / This study intends to verify how inelastic is the spending of money, with higher education in relation to the income. We found that families with higher income, spend more on that kind of education than those of lower. We observed also in Brazil, that as higher the incomes more is spent on high level education, but this correlation is inelastic, with an increase of 1,0% on the month income, carries 0,31% increase in monthly expenditure on tertiary education. In relation to the amount spent on education, the family income, we may observe that when the family income increases in certain geographic regions, a small part of it is reserved for high level education than in other regions, as we could verify. This suggests that families with high income levels, will not be affected when deciding to invest more in education to have a better quality of education compared to others. We may observe that among the brazilian regions, there are differences that often come from the number of residents and educational differences, usually in the same family. In families with higher income, we found often that part of this increase was forwarded to other activities, and this will not change so much its decision on investing in university education. It was verified that this occurs in the Southeast and South, because these locations revenue is above the national average and the number of residents per household is relatively lower. We also observed that in these regions the ratio of student is higher, confirming that they are the ones with better economic conditions and thus, they have better opportunity to invest in education. / O estudo em questão pretende verificar, o quão é inelástico o gasto com a educação de nível superior em relação à renda. Verificamos que os domicílios com maior renda há um gasto maior dos que os de menor renda. O que também foi verificado no Brasil é que, quanto maior a renda, maior é o gasto com educação de nível superior, porém esta correlação é inelástica, ou seja, com um aumento de 1,0% na renda mensal, acarreta 0,31% de aumento na despesa mensal com educação de nível superior. Quanto à proporção de gastos com educação na renda domiciliar, há evidências que com o aumento da renda em domicílios de certas Regiões Geográficas, há uma destinação de um percentual menor de sua renda para com os gastos em educação superior do que em outras regiões, conforme foi verificado. Isto leva a crer que em domicílios com um nível de renda maior, esta alteração de renda não influenciará tanto em sua decisão de investir mais em educação para ter um curso universitário de melhor qualidade de ensino. Pode-se observar que entre as regiões brasileiras, há diferenças que muitas vezes são oriundas da quantidade de moradores e diferenças educacionais, muitas vezes no próprio domicílio. Nos domicílios de maior renda, em um grande número de vezes, parte deste incremento de renda é alocada para outras atividades, pois isto não alterará em muito sua decisão relativa ao investimento no ensino superior. Foi verificado que isto ocorre nas Regiões Sudeste e Sul, pois nesses locais a renda é superior à média nacional e a quantidade de moradores por domicílio é relativamente menor. Observamos também que nestas regiões a relação de vagas por estudante é maior, corroborando que como são as regiões mais ricas, elas têm maior condição de investir na educação de nível superior.