In the modern world, Immigration has been one of the most significant as well as visible aspects of globalization. There are a growing number of peo

Literature Review

In the modern world, Immigration has been one of the most significant as well as visible aspects of globalization. There are a growing number of people moving across countries in search of better lifestyles and opportunities. Since the longest time, the key questions regarding immigration have revolved around the costs and benefits for the receiving economy and fears of the impact on labour market opportunities mainly employment and wages for the resident population has remained.

The base of immigration and migration studies started with theoretical models, gradually evolving to empirical evidence. This paper discusses both.

Theoretical

One of the earliest works of theoretical examination of the probable labour market effects, was by (George E. Johnson, 1980) whos study came at a time where the shaping up of immigration policies was gaining relevance and countering illegal immigration was also up for discussion had it be uncontrollable. The labour force in the United States was divided into two skill categories; skilled and unskilled but the focus remained on the extent to which domestic low-skilled workers would lose their job to immigrants. To investigate this, a theoretical model was designed which would estimate the impact that each immigrant would have on the employment of the domestic population, GNP and on the distribution of income. Johnson concluded that during non-recessionary periods immigration impacts the wages of low-skilled workers the most rather than employment of low-skilled native workers. He on the other hand also stated that while low-skilled wages drop, immigration increases earnings for high-skilled workers as well as the owners of capital. As far as the distributional consequences are concerned, the outlook was positive in the sense that a large domestic low skilled labour force allowed for the next generation to train for becoming highly skilled professionals. Even though the data did not prove this hypothesis, it was seen consistent with the fact that massive immigration drove up schooling and induced the native population to acquire higher average level skills which they wouldnt have otherwise. Overall, the conclusion was mixed rather than being a definitive one.

Similarly, to Johnson, (Barry R. Chiswick, 1982) in his paper discussed about many aspects of immigration into the United States labour market. Starting with the number and origin, he stated that relative to the population there was a fivefold increase seen in immigrants from the 1930s to 1970s and their origin was slowly shifting from Europe and Canada to countries in Asia, Latin America and the Caribbean. He put forth the laws for the legal entry of immigrants in the US and how they were primarily based on kinship ties, a certain reserved number of visas could also be issued for skilled based occupational preferences. Barry (1982) also detailed statistics revolving around illegal immigration that seemed to suggest an increase and looked at instances wherein workers bought in for programs would continue to live in the country based on temporary work. A unique point that he brought out as a contribution to immigrant and migrant literature was that of Productivity, based on which he classified the people entering the US as economic migrants, kinship tied migrants and refugees. In his opinion it was the economic migrants that were more self-selected for their economic success and that there was a direct relationship between occupational status and the number of years spent in the US. Consistent to Johnson (1980), he too had a mixed response to immigration and went on to suggest that immigration policies should neither have a completely open door, nor a completely closed one. They should be based on rationing visas and levels of law enforcement.

Addressing a similar issue of self- selection by Barry, (George J. Borjas, 1987) focuses on the determination of labour market quality of foreign born persons rather than the quantity. According to him immigrants in the US are not necessarily drawn from the most able and ambitious countries and for positive selection to take place there are two conditions that must be satisfied. The first being a strong and positive correlation between expectations of earnings both at home and in the US and the second, the US having a more unequal income distribution than the home country. He makes use of a theoretical framework to empirically analyse the income differential between US and the different countries people migrate from in order to understand if that motivates immigration. The result satisfies the first condition indicating a selection bias meaning immigrants to the US are not always drawn from the upper tail of their home income distribution and there exists a substantial variance in these earnings across the origin countries. Further, this variance in the quality of measures of the immigrants can be explained to a large extent by key economic and political variables in the sense that immigrants with a high income in the US relative to their measured skills come from countries with higher levels of GNP, low-income inequality and systems that are politically competitive.

Moving beyond just looking at income or wages (Rachel M. Friedberg and Jennifer Hunt, 1995) in their paper theoretically evaluate the impact of immigration on both wages and employment as labour market outcomes. They do away from the conventional way of looking at immigration as something that has only adverse effects and adopt a rather optimistic approach in the form of studying its benefits. By analysing past literature, they are able to conclude that there exists no study that provides evidence of the fact that immigration is responsible for economically significant reductions in native employment. Further, with whatever little empirical literature available at that time they establish that a ten percent increase in the fraction of immigrants at most reduce native wages by one percent even for those natives who are the closest substitutes with immigrant labour. Paving way for future research, Friedberg and Hunt state that there exists a dependency of native income growth on the human capital levels of immigrants and that empirical research has remained largely conflicted in these issues with more work needed to be done.

Fast forwarding a few years ahead (Christian Dustmann, Albrecht Glitz and Tommaso Frattini, 2008) discussed further the simplest theoretical model wherein they explained the impact immigration would have on the economic outcomes of native workers. According to them immigration could affect outcomes depending upon the difference in skill structure of natives and immigrants as well as the elasticity of capital supply. With a perfectly elastic capital supply and similar skill structure, immigration would have no impact on natives labour market outcomes and the economy would absorb the additional labour force through expansion. On the other hand, if there is a vast difference in the skill structure, then absorption of additional labour would lead to wage adjustments. These wage adjustments would lead to re-distribution and who would gain or lose would be dependent upon the skill mix of immigrants to that of the natives.

Upon the summarization of three recent empirical studies from the UK, it was seen that none of them found evidence for negative average wage effects but there existed a wage competition at the lower end of the distribution. At the same time, there were average gains to wages that could not be explained by just surplus, hence indicating towards the failure of standard equilibrium theorems in capturing all the possible mechanisms by which immigration affects labour market outcomes of natives.

Empirical Literature

While a theoretical point of view on how immigration affects labour market outcomes was a good starting point, it doesnt yield proper results in establishing a causal relationship between the two. In the case of past literature, researchers understood the importance of empirical evidence early on. On similar lines of the work of (George E. Johnson, 1980), (Joseph G. Altonji and David Card ,1989) were one of the first to empirically test the extent to which the arrival of immigrants harms or helps the less skilled natives with the help of a variation approach, of fraction of immigrants across different cities. Their paper departed from previous work in two ways, first, the theoretical model they built led to an empirical specification in the end where wage and employment outcomes of less skilled natives would vary based on the share and skill composition of immigrants in the local labour market rather than their national origin. Secondly, they allowed for supply side effects associated with possible native worker crowding out in response to lower wages as well as demand side effects associated with local population increases. Further, they also addressed the question of competition and displacement of less-skilled natives from industry-specific labour markets in the same city. Results showed a modest degree of competition as immigrants were not concentrated sufficiently in industries employing less-skilled natives to have an effect large enough on the native groups. However, there was some evidence that indicated the shifting out of less-skilled natives in high immigrant cities from immigrant intensive industries, enabling certain low-wage industries to survive in these high immigrant cities. The last section of the paper showed results from regression analysis/spatial co-relation between immigrant shares and employment/wage outcomes of natives. The co-efficient of corelation could then be interpreted to study the impact. It was seen that increased immigration had no impact on employment rates of less-skilled natives but there was a negative effect on native wages.

With vast number of papers making use of spatial correlation as their methodology, (George J. Borjas, 1999) was critical of it and pointed out two problems. The first, immigrants might not be distributed randomly across labour markets and that a positive corelation would only indicate that immigrants chose to be in areas doing well rather than measuring the extent of the complementarity between natives and immigrants. Second, the natives might respond to the immigrants entering by shifting their labour or capital to other localities until native wages and returns are equalised across areas. He then presented a simple economic model illustrating how these spatial correlations are unable to estimate parameters of interest unless they are adjusted to estimate the true wage effects of immigration. According to him, the correct specification of wage change regression is the wage change in the region regressed on net supply shock induced by immigration.

Based on an analysis of various approaches and empirical evidence of the United States from the past, including his own (Borjas,1992), The factor proportions approach (a nations actual labour supply in particular skill groups is compared to the supply it would have had in the absence of immigration, outside information on the elasticity of substitution among skill groups is then used to estimate the relative wage consequences of a supply shock), Borjas concludes by stressing upon the importance of the joint application of economic theory and econometric methods for advancement in the field.

Moving the focus from studying the impact of immigration in just the United States, (Christian Dustmann, Francesca Fabbri and Ian Preston, 2004) filled a research gap by presenting an evidence-based paper for Britain. Theoretically, they lay out circumstances in which one could expect immigration to have or not have an impact on labour market outcomes of native workers. They stated that models that assumed a limited flexibility of output mix or closedness to international trade usually predicted a long run impact of immigration on wage and employment whereas, models assuming a higher degree of flexibility of output mix and openness to trade predicted an absence of long-run effects of at least small-scale immigration on labour market outcomes. By making assumptions about the output mix, supply and skills of immigrants we can obtain a variety of outcomes that are compatible with economic theory. One such exposition is that of factor price insensitivity, where supply is inelastic, and goods are heterogenous. In such a case Dustmann, Fabbri and Preston highlight that rather than impacting wages the long run impact of immigration was felt in the output mix with the good production relatively expanding using unskilled labour. With respect to the methodology, as seen before, the dominant approach has been spatial correlations but as suggested by this paper there exist multiple problems. The first problem is the construction of a counterfactual which is not observable and hence includes assumptions that are debatable. Secondly, the variation approach used by (Joseph G. Altonji and David Card ,1989) in their paper can be problematic if immigrant shares and levels of labour market outcomes are spatially correlated due to common fixed influences. Lastly, the issue of selective settlement leads to a biased estimate as immigrants might get attracted to areas currently enjoying economic success.

This paper contributes to literature by performing an analysis of Britain as drawing conclusions from other countries is risky due to a difference in the skill mix which in the case of Britain have been similar for that of natives and immigrants. It does so by using variation in immigration to different spatial areas and then instrument this by the variation in historical settlement patterns. Findings indicate that there is little evidence of immigration having adverse effects on native outcomes and that short run and long run effects cant be differentiated from the data given. This study was the first to open opportunities for future research in this field for Britain and hence a more refined theoretical version of it came out a few years later which is discussed in the previous section.

Carrying on the investigation of the impact immigration has had on employment opportunities and earnings of the non-migrant workers in the UK, (Timothy J. Hatton and Massimiliano Tani, 2005) looked beyond the Spatial approach that concentrates on regional or local labour markets within a country, examining these effects based on different rates of immigrations and then relating them to labour market outcomes. Their aim was to enquire the possible mechanism by which the labour markets of Britain would adjust to immigration for which they put forward the case of mobility wherein the adjustment of local markets to immigration happens through immigrant inflow in one region following an outflow of the native borns to other localities, as seen in the case of US. On the other hand, an absence of mobility could be observed through the differences across localities and perfect mobility would be unobservable locally due to the effects being spread across the country. By empirically estimating the effects on net internal migration of net immigration for 11 Britain regions, this paper sheds light on inter-regional migration being an important adjustment mechanism, specially in southern England which has the largest inflow of immigrants in relation to the regional population.

An essential implication of these findings was to not undermine the impact that immigration could have on the entire economy of a country based on a smaller impact in a particular region.

Contrary to popular findings (specially 2004 dustmann) that indicate immigration as having a benign effect on labour market outcomes of Britain, (Marco manacorda, Alan Manning and Jonathan Wadsworth, 2010) suggested that their exists evidence of a change in supply of educated natives having a significant impact on their wages. They offer a solution to this puzzle by stating that in the United Kingdom, foreign born workers and native-born workers are imperfect substitutes of each other. Pooled time-series of British cross-sectional micro data on male wages is taken to empirically arrive at a conclusion defying previous ones. A large increase in immigration over the past years is seen to have had changed the wage structure, but not of the native-born workers. Results from this study, post multiple robustness checks shows a depress in earnings of previous natives in comparison to the native-born and sensitivity of recent immigrant flows to the new ones. This rise in return of wages for the native-born occurs due to a slow growth in the supply of native university workers suggesting the main negative impact of immigration is on the wages of immigrants already present in the UK. Heterogeneity is also indicated from the data, in the sense that longer-term immigrants are much closer a substitute to the native worker than the recent immigrants. Can it be added after 2004 dustmann??

With a mixed consensus on what impact immigration plays, many papers have focused on improving the way this impact can be measured. Methodologies have been constantly evolving to better suit data and to arrive at concrete conclusions. While the spatial correlation approach remains largely at play for most researchers even today, the incorporation of instruments to get more precise results has been widely popular. One such suggestion was made by (David A. jaeger, Joakim Ruist and Jan Stuhler, 2018), which moved away from the conventional shift-share instrumentation. The reason behind this is the problem of endogeneity that comes into play when only the geographical location of immigrants is looked at. Immigration might not be the only reason why people shift locations. Furthermore, making use of past patterns can also cause issues in the sense that patterns of immigration have changed due to an increase in globalisation and hence this prediction of current patterns might not work out. The idea presented in their paper is to isolate geographic variation that occurs only due to immigration and not because of any supply or demand shocks and to look at the country of origin rather than the host country.

A multi-instrumentation approach is thus put forward wherein both past and present immigration inflows are instrumented with the past settlement instrument varying only in national components. For this to work, a time period has to be chosen such that there is a vast change in the composition of people migrating to another country. By taking data for USA this paper concludes that there exists a negative but short-term impact of immigrant inflow on local wages in areas where this inflow is large. Such an approach imposes a heavy demand on data in comparison to single instrumentation and thus far has not been used in the analysis of immigration impact for the United Kingdom.

DATA AND METHODOLOGY

The Dataset

The data for undertaking the empirical investigation in this research has been extracted from the British Labour Force Survey (LFS). The LFS is a survey of households living at private addresses in the United Kingdom and aims at providing information on the UK labour market which can be further used to develop, report and evaluate labour market policies. It also contains information pertaining to migration/immigration into and within the UK.

The dataset extracted contains panel data which will then be used for analysis and deriving results in order to investigate a causal relationship between immigration and labour market outcomes (earnings and unemployment rate). Observations are taken for 5 years and between the time period 2015-2019 and are defined for twelve regions within the United Kingdom. These regions are North West, North East, South West and South East England, West and East Midlands, Wales, Scotland and Northern Ireland. The total observations sum up to (add a number) and are enough in number to give a representative sample. The table below shows the variables included in the analysis and the summary statistics for each.

Running fixed effects on this regression equation with instrumentation

Only have to use unemployment, wages, education, Gross Value added and replacing B10 (EU Migration Laws) with another instrument

4699001587500The topic is impact of immigration on labour market outcomes (which in this case are their wages and unemployment) of british workers

I will be sending the excel file with the data in panel form some of the variables are not measured for the same time period so I need help majorly with that.

Attaching the screenshot for your reference.

31282174676000The problem stands in the part where immigration is from 2014-2015 and unemployment rates are given only for 2015 i.e. data is not for the same time period. Is it possible to tun a regression in this case? Thats the main question