Economics of Immigration

Transcription

1 ECON 361: Labor Eonomis Eonomis of Immigration Why immigration beome vital in Canadian Eonomi Poliy? 1. Immigration may be ontributing to the deterioration in labor market onditions suh as inreased labor unemploymen and inreased wage polarization. 2. Pereption that immigrants pose a burden to Canadian Soial Programs that has been suffering from utbaks. 3. Changing ethni and geographi omposition of migrants may mean a hange in assimilation rates, the argument to whih when extended may mean inreased polarization between the haves and the have not. 4. Insofar as latter day migrants are beoming a large proportion of the Canadian labor fore, and that this trend is a manifestation of the globalization of labor markets, Canadian or otherwise, it would be of substantive importane to gain an understanding of its eonomi impliations. A Simple Model of the Immigration Deision Why Eonomi Agents Migrate: Differential in Wage Strutures and Geographi Mobility The simplest way to model or think about migration is to use the human apital framework we used earlier. Consider an individual or family onsidering whether to migrate from one ountry, A to B, or one state to another. Let earnings for this individual C be w t where t indexes the period and C indexes the loation. Further let the ost of migration be M. Then what these eonomi agents onsider is the following: T T A t A t B B PV = β wt β wt M = PV t= 0 < t= 0 Where PV denotes the net present value. We have as before assume that the outome is ertain. This simple model an be extended to fator unertainty in the host ountry or state, as well as eonomi outlook of the destination ountry. Then what an agent would migrate if an only if the present value of inome or outomes are greater in the host ountry than in the ountry of origin. What are some empirially testable outomes generated by this simple model? 1. An improvement in the eonomi opportunities in the destination ountry raises the net gains to migration, and hene the likelihood of migration. 2. An inrease in ost of migration, suh as inreased paper work or ost of paper work, or inreased barriers suh as seurity heks, redues the gains to migration and hene redues the likelihood of migration. Some Stylized Fats Regarding Immigration in Canada. 1. Until mid 1980s overall immigration levels flutuated onsiderably with the eonomy, and sine then, Canada admits 200,000 immigrants per year, but the numbers are nonetheless omparable and in fat slightly lower (based on per apita).

2 ECON 361: Labor Eonomis 2. Soure Countries are very different. In the mid 1960, migrants ame from the US, U.K, and Western Europe. Today, the main groups are Asians. 3. Of all migrants, about 85 perent reside in the Census Metropolitan Areas (CMA), ompared to 55.2% of natives. What an poliy makers ontrol? 1. The numbers admitted. 2. Who gets admitted. Historially, immigration into Canada was open to anyone, with the exeption of Chinese. This trend was urtailed in the mid 1960s, both in numbers and type. But what really affets these deisions. We venture some here. 1. The skills that migrants bring that may be urrently in short supply. This group of migrants enter on what they an bring to Canada, and is typially referred to as the assessed lass, independent immigrants. The manner in whih this works today (sine 1967) is through the point system. This serve as a guide to bureaurats in examining the likelihood of suessful integration into Canadian eonomy and ulture. 2. Sine urrent migrants will be future natives, poliies would aim to allow family reunifiation. Typially referred to as non-assessed, or family lass. 3. Humanitarian onerns suh as refugees from war torn regions, and is referred to as refugee lass. This is also a non-assessed lass. What are some of the onerns reated by suh a immigration system? 1. What is the impat of immigrants on the Canadian labor and onsumer markets? If immigrants have not adverse impat on the native eonomy, then poliy makers need only fous on humanitarian onerns. However, if depending on who enters, impat may be differen poliy makers have to evaluate who should be permitted to enter into Canada. 2. How do immigrants perform in Canada? We need to know whih qualities would best enourage assimilation into the general populae, and onsequently be vital elements within the Canadian eonomy. How do Immigrants Affet the Canadian Labor Market? The simplest way to onsider this question is through the standard demand and supply analysis. 1. Inrease in new immigrants inreases the population of potential workers, and inreases the labor supply, hene suppressing the wages, and if eonomy is operating at full employmen this would result in inreased unemployment and beome a burden to soial programs.

3 ECON 361: Labor Eonomis Supply of Labor Inreases due to inreased immigrants Wages fall 2. However, if the assessed lass is admitted as a result of labor shortages, the above naïve analysis annot be true. In fat if entry is based on shortage, so that prior to entry the labor market is in disequilibrium, immigrants then relieve shortage without any adverse effet. 3. Immigrants may bring about an inrease in demand for goods, raising the derived demand for labor, hene aneling out the possibility of wages being surpressed. 4. Change trade patterns due to information that immigrants bring to Canada about their soure ountry. 5. Immigration may be a substitute for importation of goods. 6. Immigrants may also invest in Canadian ompanies, or start their own businesses, thereby raising employment in the Canadian eonomy. Estimation Problems: 1. Impat of immigrants are dependent on the relevant setors in whih they enter, and hene is diffiult to segment to entire labor market to disern the real impat. 2. Lak of ounter-fatual, i.e. we do not know how the labor market would have been had there be no immigrants. 3. The timing of entry by immigrants is endogenous, i.e. dependent on the going eonomi onditions. This then impedes our ability to understand the impat of immigration. Possible Solutions: 1. Measure the degree of substitutability between native and immigrant labor. This is done by estimating the aggregate prodution funtion of the eonomy, using ross-industry, and ross-oupation mixes of immigrants to identify substitutability. If the latter is omplementary to the former, the greater the proportion of immigrants, the greater would prodution or output be (impat that is greater than 1), while if substitutable, the greater the proportion of immigrant

4 ECON 361: Labor Eonomis would imply no hange to the aggregate output. It has been found that this immigrants has little impat exept in low skilled labor markets. 2. Using time series, examine the hanges in unemployment rates as a result of immigration. It has been found that immigration has very small adverse impat. Some aveats here should be noted. This mild adverse impat may simply reflet adjustment time of immigrants themselves so that in the long run, immigrants has no impat of unemployment. However, we annot separate the unemployment rate into immigrant and native labor. Further, although there is no aggregate impat on unemploymen we annot generalize this observation into the miro impat on the various industries and omprise the labor market. 3. Sine ity of hoie vary among immigrants, we ould exploit this variation to examine the net impat immigrants have on the destination. If all ities have the same eonomi irumstane, the adverse impat of immigration would suppress wages in the ity with higher reent immigrant onentrations as ompared to another with lower immigrant onentrations. However, not all ities are the same, and insomuh as they are differen it would affet our preditions. That is the unobservable differenes that attrat immigrants to different ities may be orrelated with the proportion of immigrants in a ity (An example is how osmopolitan a ity already is, whether the ity has a substantial ethniity of whih the immigrant belongs). One solution is to examine the hange in immigration rates, and its impat on hanges in wages and other outomes instead. The idea is that this proedure would anel out all intervening permanent harateristis of the ity. Take for example the following: Let an immigrant hoose a ity to immigrate to, be dependent on the level of disrimination that a immigrant have to fae. This is an unobservable quality of a ity, and will be inluded in the error term sine we have not aounted for it. The manner in whih disrimination affets the wage is through firms hiring on the basis of itizenship status. Then the regression of wages without onsidering this impat or onern would be w = a + a prop + AX + u w 1 = a a 2 im, prop im, + AX + ( e + dis ) However, if we run the regression on differenes over time, the following is obtained w w = a + a prop prop + A X X + e e ( ) ( ) ( ) t 1, 1 2 im, im, t 1, t 1, t t 1 Note now the disrimination term is anelled ou and we an estimate the oeffiient of the impat of immigration on wages. Nonetheless, the issue of seletion into the various ities remain. Further the analysis still ignores the need for a ounterfatual. What would have happened to Toronto without immigrants. It has been found that immigration seems to have little impat on the native s eonomi outome if any, and is at most onentrated in low skilled labor markets. (Note that there is also the problem that eonomi irumstane of ities attrat immigrants, and not that immigrants bring good fortune to a ity, that is there is still the problem of simultaneity.) 4. Use of natural experiments to measure the impat of immigration, suh as flood of refugees due to exogenous irumstane. The results as before shows little

5 ECON 361: Labor Eonomis impat of immigration. (However, what is the eonomi and soial status of these individuals or families fored to immigrate.) Eonomi Performane of Immigrants We now examine how well immigrants assimilate into their destination environments. 1. Immediately upon arrival, immigrants fae unemployment as they go about their job searh. Adjustment may reveal itself through; a. Lower hours of work. b. Lower wages beause redentials not transferable, and/or language, and/or lak of knowledge of loal labor marke and/or differenes in skill requirements. 2. With time, it is expeted that immigrants wages would tend towards that of natives. If immigrants are positively seleted, that is if Canada s immigration poliy is effetive in seleting immigrants best suited for the needs of the labor marke their wages may eventually exeed that of natives. 3. If immigrants have to fae disrimination, passage of time may mean a dissipation of these effets, and hene lead to a rise in wages. Based on the above arguments, the earning profile over time for an immigrant labor would like the following, Earnings Immigrant Native Entry Effet 0 T 40 Years sine Migration What are some of the key features of the above hypothetial assimilation profile? 1. Comparing two individuals who differ only on one aspe their state of residene at birth. The above profile essentially says that based on the above points, an immigrant would suffer an entry effet on arrival when ompared to a native of the same age and on other observable and unobservable traits.

6 ECON 361: Labor Eonomis 2. With time, or as years sine immigration inreases, our arguments suggests that these immigrants would eventually assimilate, and obtain the same wages. 3. Further if there is positive seletion, the wages of immigrants may even ontinue to grow at a faster rate than native, hene exeeding the native s earning as time progresses. 4. The time to assimilation, T, is dependent on how large the initial entry effets are, and the rate of assimilation. The faster the assimilation rate, the steeper the immigrant s assimilation profile, and the smaller T is. Similarly, the greater the entry effet is, the longer it would take for assimilation to our, and the larger T would be. Estimation and Empirial Problems 1. Using Cross Setional Analysis on Census data (using ensus from different years, i.e. pooled ross setion), we an ompare immigrants and natives who possess similar harateristis. By omparing reent immigrants to natives of the same age and harateristis we an hene estimate the entry effet. By omparing reent immigrants with immigrants who has been in the ountry for a longer period of time, we an infer the rate of assimilation of immigrants. However, there are some problems with this approah. a. Cohort Effets: This assumes that eah ohort faes the same entry effets, and the same assimilation rates. If violated, we will not be able to disern the true entry effets, and assimilation rates. b. Self Seletion: We have to ontend with the possibility that in any ensus period, the remaining immigrants who remains in the post initial immigration period is representative of those who are just arriving. What if this is not true, suh as when immigrants who leave Canada are systematially the better ones (upper tails in the distribution of immigrant type) or the lower ones (lower tails)? It would be diffiult to solve this problem even with a quasi panel where we ompare immigrants from the same immigration ohor but aross different ensus. 2. Use Panel Data and trak immigrants. Some results that were found: 1. Meng (1987), and Abbot & Beah (1993) found using the Job Mobility Survey (JMS) that entry effets was about 15%, with immigrants taking 0 to 14 years to ath up with the natives. The latter paper also found that immigrants enjoyed a higher rate of return to experiene than native, that is with time, earnings of the former superseded the latter. 2. However, Baker & Benjamin (1994), and Bloom, Grenier & Gunderson (1995) using ensus data between 1971 through to 1986 found that a substantial inrease in entry effets, and reported negligible assimilation rates. In fat the former paper found that the sharp inrease in entry effets is orrelated with the soure ountry among latter day immigrants. It should be noted that Grant (1999) using to ensus years of 1986 and 1991 found signifiant assimilation rates (17%), and also found that entry effets has stopped inreasing. She argues that the hanges were likely due to better eonomi irumstane in the late 1980s. Why the differene?

7 ECON 361: Labor Eonomis MDonald and Worswik (1998) suggests that it is diffiult to identify the true entry effets and assimilation rates, and found onlusions to be very sensitive to the researhers hoie of ensus years, prinipally due to the differential in points along a business yle in whih the surveys were drawn. 3. Beah & Worswik (1993) using the JMS found that although immigrant women enjoy earnings premium over native female labor, but lower assimilation rates. Bloom, Grenier and Gunderson (1995) found however little entry premium, lower entry effets as ompared to men, and no assimilation as well. The argument for this is based on the Family Investment Hypothesis where women enter the labor market and take on jobs with low prospets as a result of borrowing onstraints due to their immigrant status. 4. Baker and Benjamin (1995a, & b) and Crossley, MDonald, and Worswik (2001) found that despite poor eonomi outomes, immigrants were less ikely to use unemployment insurane and soial assistane. Akbari (1995) using 1991 ensus found that taxes paid out by immigrants far outweigh transfer reeipts. Hene arguments that immigrants who fail to assimilate represents a burden to Canadian taxpayers does not seem to hold. 5. On another level, we would also like to know if the Point System adopted in Canada were better able in hoosing immigrants who were more likely to adapt and assimilate, with lower entry effets. Green & Green (1995), and Wright & Maxim (1993) found that the poliy has indeed been suessful. That is the assessed lass of immigrants fare better than the non-assessed. That immigrants from developing ountries fare the worse.