. 1 Give Us Your Wired and Skilled: Measuring the Impact of Immigration Policy on Employers and Shareholders Carl Shu-Ming Lin0F * Department of Economics, Rutgers University June, 2010 Abstract This paper attempts to link financial economics theory to the area of labor economics (particularly in migration) and political economy (immigration policy). By using event study analysis and the market model, I measure the impact of immigration policy on the welfare of employers and shareholders, in particular on those industries which hire large number of skilled immigrants. Due to the high demand of skilled workers (especially in high-tech industries), the American Competitiveness and Workforce Improvement Act (ACWIA) of 1998 greatly raised (nearly doubled) the available number of H-1B visas for skilled foreign workers for the first time since 1990. Hence, this paper focuses on this bill and analyzes whether it increases shareholders’ profit. And if yes, by how much. The empirical results show that top H-1B user industries benefit from the ACWIA of 1998. Employers and shareholders of those industries enjoy significant and positive returns after the bill was passed. Shareholders of high-tech industries (top users of H-1B visa, 80% of total) such as Computers and related equipment, Computer and data processing services gain an average 16.78% and 18.65% cumulative excess return in a month after the Act was passed, respectively. Non-high-tech industries (remaining balance of H-1B visa) such as Security, commodity brokerage, and investment companies has a relatively lower cumulative excess return (average 2.37%). JEL Classifications: J61, K31, G12 Keywords: Skilled immigrants, Immigration Policy, Employers, Shareholders, Market model, Event Study, H-1B visa. Correspondence: Carl Lin, Economics Department, Rutgers University, 75 Hamilton St., New Brunswick NJ 08901-1248 USA. Email: U[email protected]U Tel : +1 202-320-7266. Fax : +1 732-932-7416. * Ph.D. student in Economics, Rutgers University. I thank Orley Ashenfelter, Ira Gang, John Landon-Lane, Carolyn Meohling, Anne Morrison Piehl and Hilary Sigman for useful comments and extended discussions.
87
Embed
Give Us Your Wired and Skilled: Measuring the Impact of ... · Give Us Your Wired and Skilled: Measuring the Impact of Immigration Policy on Employers and Shareholders Carl Shu-Ming
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
.
1
Give Us Your Wired and Skilled:
Measuring the Impact of Immigration Policy on Employers and
Shareholders
Carl Shu-Ming Lin0F
*
Department of Economics, Rutgers University
June, 2010
Abstract
This paper attempts to link financial economics theory to the area of labor economics (particularly in
migration) and political economy (immigration policy). By using event study analysis and the market
model, I measure the impact of immigration policy on the welfare of employers and shareholders, in
particular on those industries which hire large number of skilled immigrants. Due to the high demand
of skilled workers (especially in high-tech industries), the American Competitiveness and Workforce
Improvement Act (ACWIA) of 1998 greatly raised (nearly doubled) the available number of H-1B visas
for skilled foreign workers for the first time since 1990. Hence, this paper focuses on this bill and
analyzes whether it increases shareholders’ profit. And if yes, by how much.
The empirical results show that top H-1B user industries benefit from the ACWIA of 1998.
Employers and shareholders of those industries enjoy significant and positive returns after the bill was
passed. Shareholders of high-tech industries (top users of H-1B visa, 80% of total) such as Computers
and related equipment, Computer and data processing services gain an average 16.78% and 18.65%
cumulative excess return in a month after the Act was passed, respectively. Non-high-tech industries
(remaining balance of H-1B visa) such as Security, commodity brokerage, and investment companies
has a relatively lower cumulative excess return (average 2.37%).
Correspondence: Carl Lin, Economics Department, Rutgers University, 75 Hamilton
St., New Brunswick NJ 08901-1248 USA. Email: [email protected] U Tel : +1
202-320-7266. Fax : +1 732-932-7416.
* Ph.D. student in Economics, Rutgers University. I thank Orley Ashenfelter, Ira Gang, John Landon-Lane, Carolyn Meohling, Anne Morrison Piehl and Hilary Sigman for useful comments and extended discussions.
Most research that has been conducted on the impact of the immigration has
focused on the consequences for workers (such as employment, wages, earnings,
etc.). Less is known about the impact on employers. We lack answers to such
questions as: How much do immigrants increase or reduce employers’ profit? Which
employers are most likely to gain (suffer) increased (reduced) profits as a result of
immigration? Employers and their representatives usually support immigration;
however, employees often strongly oppose immigration. Nevertheless, we have little
evidence to assess the quantitative effect of immigration on employers’ profit.
An increase of skilled (unskilled) immigrants will raise the costs of business for
employers of skilled (unskilled) workers. On the other hand, an increase of skilled
(unskilled) immigrants may help to generate greater profits for employers by
producing more goods and services. Although employers and their representatives
often support to lift restriction on immigration, however, we still lack the evidence
which answers whether hiring more skilled (unskilled) immigrants may generate
greater profits for employers and shareholders.
This paper attempts to apply financial economics theory to the area of labor
economics (particularly in migration) and political economy (immigration policy). By
using event study analysis and the market model, I measure the impact of
immigration policy on the return of employers and shareholders, in particular on
those industries which hire large number of skilled immigrants.
Applying methodologies from finance theory on the area of labor economics can
be traced back to 1980s. Ruback and Zimmerman (1984) show that successful union
elections result in a 3.8% decline in shareholder equity of those organized firms.
Becker and Olson (1986) uses event study to analyze the impact of strikes on
shareholder equity. They show that from 1962-82, strikes substantially affect
.
3
shareholder equity as measured by the change in stock prices associated with strikes.
Over that period the average strike involving 1,000 or more workers resulted in a
4.1% drop in shareholder equity, representing a decline of $72-87 million in 1980
dollars. Becker and Olson (1989) examines union-nonunion differences in the
allocation of both firm profits and business risk to employees and shareholders.
Using a sample of more than 1,000 large private sector firms, they find that over the
period 1970-81 shareholders in unionized firms assumed less of the firm’s business
risk than shareholders in nonunion firms. In addition, risk-adjusted returns to
shareholders were lower in unionized firms than in nonunion firms. Card and Krueger
(1995) analyze how the change of the minimum wage affects low-wage employers’
profits. They focus on a large sample of publicly traded firms and find that a increase
in the minimum wage may have had a small negative effect on the value of such
firms-on the order of 1 or 2%. However, after adjusting for overall market returns,
their results provide mixed evidence that the value of these firm changes in response
to legislative maneuvering on the minimum wage.
I begin by describing estimation strategy in section 2. In section 3, I describe the
background on the H-1B program. Section 4 contains data and descriptive statistics
and section 5 provides empirical results. Finally, section 6 is the concluding remarks.
II. Empirical Methodology
A. Measuring the Welfare of Employers and Shareholders
To measure the impact of immigration policy on the welfare of employers and
shareholders, the first question is: what measure can be used to estimate their
welfare? A central assumption in the finance literature is that because capital
markets are efficient, the prices of capital assets are unbiased estimates of the
present value of future profit streams generated by those assets (Fama, Fisher,
.
4
Jensen and Roll (1969), Schwert (1981)).
Since a firm can be viewed as a bundle of capital assets, firm value or the present
value of the shareholders’ claim to this profit stream is a function of the expected
future cash flow and the variance in this cash flow. The firm’s economic profit at time
t to employers and shareholders is simply the price of an individual share of
common stock multiplied by the number of shares outstanding. If the stock market is
efficient, changes in stock prices (returns) can be interpreted as an estimate of the
change in the value of the firm caused by new information regarding the future
profitability of the firm (Becker and Olson (1986)). Hence, shareholder returns over a
given period are measured as the change in common stock prices during that period
plus dividends paid. 1F
1
Therefore, this paper uses the event study analysis and the market model to
estimate employers’ normal and abnormal returns 2F
2 under the impact of immigration
legislation. Figure 1 illustrates the time line of the event study. By defining 0t as
the event day (such as the day that the bill is passed), 1 1t T to 2T represents
the event window, and 0 1t T to 1t T is the estimation window which is used
to estimate the normal performance return of a firm.
The length of estimation window must be chosen first. Typically, 255 days is
selected to correspond approximately to the number of trading days in a calendar
year. For event window, 20 days (10 days before/after the event day), 30 days (15
days before/after the event day) and 60 days (30 days before/after the event day) are
1 Note that this paper only focuses on economic profits rather than accounting profits. Economic profits are the stream of net cash flows that accrue to shareholders, as owners of the firm’s assets, and they represent revenues minus operating costs and new investments. Shareholders’ wealth is simply the present value of these net cash flows. Strictly speaking, firm value should also include the value of debt as well. This paper focuses, however, on the interests of shareholders since the claims of bondholders are largely fixed. 2 In some research abnormal returns are called excess returns.
.
5
usually used to see the pattern of abnormal returns before/after the event day. In
order to fully understand the impact before and after the immigration legislation,
this paper chooses 60 days event window since 20 days or 30 days window may not
be enough to catch the pattern of abnormal return that is caused by the immigration
legislation. The post-event window estimation can also be obtained if needed.
Figure 1 Time Line for an Event Study
Note: Estimation window (T0 – T1)= 255 days corresponds approximately to the number of trading days in a
calendar year. Event widow (T1 – T2)= 60 days plus an event day t=0.
In sum, the abnormal return over the event window can be interpreted as a
measure of the impact of the event on the value of the firm. Thus, the methodology
implicitly assumes that the event is exogenous with respect to the change in market
value of the security. In other words, the revision in value of the firm is caused by the
event which can be viewed as the change in the welfare of employers and
shareholders.
B. Stock Market Evaluation – The Market Model
In examining the impact of skilled immigrants on shareholders, the effect of overall
market factors can be removed by estimating a standard market model. Formally, a
daily return market model can be expressed as:
it i i mt itR R (1)
2[ ] 0 [ ] ,iit itE Var
.
6
where itR is the return on the common stock of firm i on day t , adjusted for
stock splits and dividends; mtR is the return on the value-weighted
NYSE/AMEX/NASDAQ index on day t ; i and i are regression coefficients; and
it is an error term of firm i on day t .
Under general conditions and assumptions which asset returns are jointly
multivariate normally distributed with mean and covariance matrix for all t ,
ordinary least squares (OLS) is a consistent and efficient estimation procedure for the
market model parameters.
Estimated abnormal return ( AR ), also known as prediction errors, can be
calculated for each firm i for each day t in the analysis period by
ˆ
ˆˆ( )
it it
it i i mt
AR
R R
(2)
where itAR is the abnormal return of firm i on day t and ˆi , ˆ
i are estimates
of i and i .
The abnormal returns ( AR ) are estimates of the abnormal returns to the
stockholders of the sample of firms on each trading day. Mean abnormal returns
( AR ) across all firms can be calculated for each day in the analysis period. Hence,
the mean abnormal returns of an industry can be easily obtained. These averages are
then accumulated to provide a series of cumulative mean abnormal returns (CAR )
around each event. That is,
2
1
1 2( , )t
t
t t
CAR t t AR
(3)
In order to draw overall inferences for the event of interest, the abnormal return
observations must be aggregated. The aggregation is along two dimensions –
.
7
through time and across firms.
To proceed, one first obtain the normal performance return (during estimation
window 255 days) of firm i by running the market model equation (1). Second,
abnormal return of firm i on trading day t is calculated using equation (2). Third,
mean abnormal return of each industry on each trading day can be obtained by
averaging the abnormal returns of all firms in that industry. Finally, by accumulating
the mean abnormal returns (using equation (3)), one can tell the cumulative effect of
the immigration policy on the return of employers and shareholders.
C. Hypothesis Testing
After obtaining the mean abnormal returns and cumulative mean abnormal returns
of each industry which we are interested, it is important to know whether these
abnormal returns are statistically significant from zero. Hence, an intuitive method is
to test the null hypothesis that the mean abnormal performance is equal to zero.
That is, under the null hypothesis, 0H , for the event window sample abnormal
returns (prediction errors) is
*ˆ (0, V )i i (4)
Equation (4) gives the distribution for any single abnormal return observation. Vi is
the variance-covariance matrix of the estimated abnormal return *
i .
In literature, many tests have been developed to check the abnormal return
performance in a event study. Patell (1976) first uses standard normal test to check
abnormal returns assuming cross-sectional independence. To correct the serial
correlation of the abnormal returns, Mikkelson and Partch (1988) uses the corrected
version of Patell test. The corrected test accounts for the fact that within the window,
the abnormal returns for each stock are serial correlated. Boehmer, Musumeci and
.
8
Poulsen (1991) introduce an empirical cross-sectional variance adjustment in place of
the analytical variance of the total standardized prediction error.
Brown and Warner (1980) introduce a "crude dependence adjustment" test which
is also known as the time-series standard deviation test. Unlike the standard normal
return test, the time series standard deviation test uses a single variance estimate for
the entire portfolio. Therefore, the time series standard deviation test does not take
account of unequal return variances across securities. On the other hand, it avoids
the potential problem of cross-sectional correlation of security returns.
Besides the parametric tests described above, nonparametric tests can be used to
supplement the validity of the results. This paper performs the general sign test. The
idea of the nonparametric general sign test is: for each trading day in the event
period, and for each window, the number of securities with positive and negative
mean abnormal returns can be reported. The null hypothesis for the generalized sign
test is that the fraction of positive returns is the same as in the estimation period.
The empirical results of market model and hypothesis testing are reported in section
V.
III. Background on the H-1B 3F
3 Program
The H-1B program was established in 1990 4F
4 to permit skilled foreigners to work in
the United States. The program grew out of the H-1 visa program, which was created
during the early 1950s to allow firms to hire into temporary jobs skilled foreign
workers coming to the United States on a temporary basis. Beginning in 1970,
3 The H-1B visa is a category of employment-based nonimmigrant visas that allows skilled aliens in certain “specialty occupations” to work in the United States. A potential U.S. employer of an eligible foreign worker must file an H-1B visa petition with the Immigration and Naturalization Service (INS) of the Department of Justice and also a labor condition application (LCA) with the Employment of Training Administration (ETA) of the Department of Labor. 4 President George H.W. Bush's signing of the "The Immigration Act of 1990" is often considered the day H-1B was born.
.
9
employers were allowed to hire foreigners for permanent positions, and the number
of visas issued increased as the U.S. economy boomed during the 1980s.
Under the 1990 Act visas for employment-based immigrants rose to 140,000 from
the 58,000 cap established in 1976. The 1990 Act set an annual cap of 65,000
nonimmigrants entering the U.S. under H-1B visas. H-1B workers were given a 3 year
visa with a possible extension for a total of six years. It specified that H-1B workers
must hold at least a bachelor’s degree or its equivalent in their specialty field. The
Act also required employers to pay H-1B workers the prevailing wage. In addition, the
1990 Act created three other new visa categories for skilled temporary workers–the
H-1A visa for nurses and O and P visas for prominent scientists, educators, artists,
athletes and entertainers.5F
5
Figure 2 describes the annual H-1B visa issuance cap and the actual H-1B visa
issuances since 1990. The figure also shows the H-1B visa population estimated by
Lowell (2000). It is very clear that in 1998, which the American Competitiveness and
Workforce Improvement Act (ACWIA) of 1998 was passed, the actual H-1B visa
issuances and especially the estimated population started to grow in an amazing
speed. Hence, the ACWIA of 1998 is a good candidate to study the impact of the
increased skilled immigrants (in terms of H-1B visas) on the return of employers and
shareholders. Next, the background and legislation history of the ACWIA will be
discussed.
5 This part is drawn heavily from Zavodny (2003).
.
10
Figure 2 H-1B Visas and Population Estimates
Source: Kerr, William R. and Lincoln, William Fabius, The Supply Side of Innovation: H-1B Visa Reforms and US Ethnic Invention
(December 16, 2008). Harvard Business School Entrepreneurial Management Working Paper No. 09-005.
The American Competitiveness and Workforce Improvement Act (ACWIA) of 1998
During the 1990s, the booming economy, low unemployment, and a shortage of
skilled domestic workers have dramatically increased U.S. employers’ demand for
skilled foreign workers. This trend is especially strong in the information technology
(IT) and computer industries. For many years, the U.S. high-tech industry has become
the dominant participant in the H-1B visa programs. Prior to 1998, U.S. Congress had
capped the annual quota of new H-1B at 65,000. Because of this limitation, the
existing H-1B visa program could no longer meet the high-tech industry’s voracious
demand for foreign skilled workers. Since 1997, H-1B visas have been oversubscribed:
the number of H-1B admissions reached the statutory cap of 65,000 before the end
of each fiscal year, and employers petitioning late in the year would be required to
wait another year for the admission of approved workers.
Unable to fulfill the unprecedented needs for skilled workers under the existing
.
11
H-1B visa program, the high-tech industry actively lobbied Congress to raise the
annual cap on the number of H-1B visas granted to immigrants. However, the
congressional effort to raise the annual H-1B visa cap met vigorous opposition from a
vocal minority in Congress, labor unions, and the White House. After months of
wrangling, the White House and congressional supporters of the new H-1B bill finally
reached a compromise in the fall of 1998. On October 21 of that year, President
Clinton signed into law the compromise bill, the American Competitiveness and
Workforce Improvement Act of 1998. The new H-1B visa law nearly doubles the
available number of H-1B visas over the next three years. The law also addresses
concerns over the potentially adverse impact of the H-1B visa program on the
domestic workforce and potential abuses of the program by H-1B employers.
Legislation History of the ACWIA
In early 1998, Republican Senator Spencer Abraham of Michigan sponsored
legislation addressing the issue of the annual H-1B visa cap and the needs of the
high-technology labor market; the Senate debated the matter in early 1998. The
Senate, with little opposition, passed the American Competitiveness Act raising the
annual cap on H-1B visas on May 18, 1998. However, the attempt to raise the H-1B
visa cap met strong opposition in the House of Representatives from traditionally
pro-labor Democrats and anti-immigration Republicans.6F
6 These legislators received
the backing of labor unions 7F
7 and professional engineering organizations such as the
6 Patrick Buchanan, Commentary, Sellout of High-Tech Jobs, Washington Times, August 19, 1998, at A17 (criticizing the H-1B visa program for transforming the American workplace into the “Asian environment,” and the Silicon Valley companies for failing to “Americanize” their labor force); Spencer Abraham and David McIntosh, Commentary, Why America Needs Temporary Foreign Workers, Washington Times, September 1, 1998, at A16 (“On this issue [of H-1B visas], Pat Buchanan... [is] wrong, and America's innovators are right.”); William Branigin, House Sets Aside Bill to Allow Hiring of More Foreign Workers: Measure Sought by High-Tech Firms Had Split GOP, Washington Post, August 1, 1998, at A2 (discussing the split among Republicans on the issue of raising the H-1B visa cap). 7 William J. Holstein, Give Us Your Wired, Your Highly Skilled: Tech Firms Are Winning the Battle of the Visas, U.S. News & World Rep., October 5, 1998, at 53 (reporting the demands of labor organizations
.
12
Institute of Electrical and Electronics Engineers-USA (IEEE-USA).8F
8 The opposition to
the proposal to raise the H-1B visa cap transcended traditional party lines, forming
an odd coalition of liberal, pro-labor Democrats and conservative, anti-immigration
Republicans.
Under pressure from labor unions and pro-labor Democrats, the White House
initially opposed the new H-1B visa bill due to concerns over the perceived
inadequacy of the job-protection provisions in the original bill. As the House of
Representatives, prepared to consider the bill before the August recess “the White
House issued a public veto threat and listed... changes it was seeking to the bill.”
After months of wrangling and intense negotiations, the White House and the
congressional supporters of the bill finally reached a compromise on September 23,
1998, in which they agreed to raise the H-1B visa cap while including additional
protective measures for American workers. The House passed the new compromise
H-1B visa bill the next day. However, the bill faced an unexpected sudden death in
the Senate on October 8, when a small number of senators led by Democrat Tom
Harkin of Iowa blocked the vote. 9F
9 After a skillful legislative maneuver by its
supporters, the H-1B visa bill made a remarkable, eleventh-hour comeback as part of
the omnibus appropriations bill on October 16.10F
10 On October 21, 1998, President
like the Communications Workers of America and the AFL-CIO that “Americans displaced by global competition or downsizings ought to have first priority in taking the high-paying jobs”). 8 John R. Reinert, Commentary, Trojan Horse in the Free Labor Market?, Washington Times, Sept. 26, 1998, at C2 (asserting that the H-1B visa program hurts U.S. engineers); Zitner, supra note 16, at C1 (quoting IEEE-USA president John Reinert as stating that “[the evidence doesn't suggest that there is a labor shortage, and there is no need to increase the number of visas”). According to IEEE-USA, a report by an outplacement firm showed that high-tech industries have laid off 143,000 workers in 1998, more than any other sector of the economy; Robert MacMillan, H-1B Visa Bill Ready for Passage, Newsbytes, October 8, 1998, available at LEXIS, News Library, Wire Service Stories File (reporting IEEE-USA president-elect Paul Kostek's argument that "it's bizarre policy to give the industries laying off the most US workers special access to an expanded foreign guest-worker program"). 9 Bill to Bring Technology Workers to U.S. Dies, New York Times, Oct. 10, 1998, at C2 (reporting the 11th-hour death of the H-1B visa bill); Ashley Dunn, Plan to Increase High-Tech Work Visas Dies in Senate, L.A. Times, October 10, 1998, at C1. 10 Mark Leibovich, High Tech Is King of the Hill: Rash of Legislative Wins Has Industry Celebrating, Washington Post, October 16, 1998, at F1 (“Technology lobbyists, executives and congressional
.
13
Clinton signed the controversial compromise H-1B visa bill into law: the American
Competitiveness and Workforce Improvement Act of 1998.
supporters managed to attach the bill to the broader budget package.”).
.
14
IV. Data and Descriptive Statistics
Since the U.S. Citizenship and Immigration Services (USCIS) 11F
12 does not provide the
detailed statistics of H-1B visa of Fiscal Year 1998 and 1999 12F
13, the alternative way to
analyze the skilled immigrants is to use the data from the 1% 2001 to 2008 American
Community Survey data and 5% 1980, 1900, 2000 Integrated Public Use Microdata
Series USA. Nevertheless, it is still important to understand the trend and statistics of
skilled immigrants in the United States.
A person is classified as an immigrant if he or she was born in a foreign country.
The term foreign born refers to people residing in the United States at the time of
the census who were not US citizens at birth. The foreign-born population includes
naturalized citizens, lawful permanent immigrants, refugees and asylees, legal
nonimmigrants (including those on student, work, or other temporary visas), and
persons residing in the country without authorization. I restrict the analysis to
individuals who age 25 to 64, not self-employed or worked without pay and did not
reside in group quarters. Skilled immigrants are defined as those who have 13 or
greater years of schooling. Person weight is used throughout the paper.
Source of Region
Table 1-1 and 1-2 summarize the source of region of skilled immigrants. From 1980
to 2000, the largest number and proportion of skilled immigrants come from Europe.
In 1980, 31.84% of immigrants were from Europe; then it decreased to 24.39% in
1990 and to 22.02% in 2000. Despite the decreasing trend, Europe is still the largest
source region of skilled immigrants in the last three decades. Other important source
of regions of skilled immigrants are East Asia (average 12.15%), Southeast Asia
12 Formerly the Immigration & Naturalization Service (INS). 13 The ACWIA of 1998 requires that information about successful H-1B visa applications be submitted each year to Congress. The first report under this requirement was submitted to Congress for those approved for H-1B status in Fiscal Year 2000.
.
15
(average 13.71%) and India/Southwest Asia (average 8.71%). Note that the trend of
the three regions is increasing over time.
Occupation
Table 2-1 and 2-2 show the seven general occupations of skilled immigrants in the
United States. In table 2-1, most skilled immigrants work as managerial and
professional specialty (average 44.25%) from 1980 to 2000. The second most popular
occupation is technical, sales and administrative support (average 31.05%).
Besides the seven general occupations, we may be interested in the detailed
occupations. There are approximately 900 occupations which can be identified from
the survey data. Table 4-1 to table 4-2 summarize the detailed occupations for skilled
immigrants. In table 4-1, managers and administrators has the largest number of
skilled immigrants in 1980, 1990 and 2000. Registered nurses, salespersons
occupation also has many skilled immigrants in 1980 and 1990. Note that in 2000,
they are replaced by computer software developers and computer system analysts. If
by percentage, physicians, physical scientists and engineers are the highest
percentage occupations among skilled immigrants. Note that physicians averages
24.17% from 1980 to 2000. Physical scientists is even as high as 40% in 2000.
Industry
For the purpose of understanding the impact of skilled immigrants on the welfare
of employers, knowing which industries hire the largest number and highest
percentage of skilled immigrant is particularly important. Table 3-1 summarizes the
thirteen general industries that immigrants are likely to work in. The industry that
has the largest number of skilled immigrants is professional and related services
(average 32.8% from 1980 to 2000). The second and the third largest are
manufacturing and retalesale trade. Manufacturing averages 18.15% and the trend is
decreasing over time. On the other hand, retalesale trade averages 10.88% and the
.
16
trend is increasing.
Since this paper tries to analyze the impact of immigrants on the welfare of
employers and shareholders, the thirteen general industries listed above are too
broad. Hence, it is necessary to narrow the thirteen general industries down to more
detailed ones. Table 5-1 summarizes the detailed industries of skilled immigrants.
Table 5-1 reports the top 10 (by number and by percentage) detailed industries
which hire skilled immigrants. The table shows an important fact that hospitals,
colleges and universities, elementary and secondary schools have the largest number
of skilled immigrants from 1980 to 2000. If we look at percentage, in 1980,
engineering, architectural, and surveying services, colleges and universities have the
highest percentage of skilled immigrants. Starting 1990, computer related industries
(computers and related equipment, computer and data processing services) and
research, development, and testing services gradually become the industries which
hire high percentage of skilled immigrants. The importance of skilled immigrants in
the computer related industries has been growing over time. Take computers and
related equipment industry for example, it was 6.95% in 1980 then increased to
11.67% in 1990 and 19.19% in 2000. The other computer related industry, computer
and data processing services, also has 18.57% skilled immigrants among all
employees.
In sum, skilled immigrants are likely to work in computer related, research and
development industries. This confirms the same fact from the occupation section
above.
V. Empirical Results
This paper focuses on the impact of skilled immigrants on the returns of employers
and shareholders during the legislation process of ACWIA of 1998 since the act
.
17
nearly doubled the H-1B visas for skilled immigrants. Lowell and Christian (2000)
reports that fully 80% of the top H-1B users in 1999 are in the IT industries, the
balance being non-IT. Among the non-IT companies there is no large, single sector
but there are clear lines of business: 7% of the top H-1B firms are in
business/management consulting, another 4% are in executive/temporary
placement services, and nearly 6% are in accounting/ engineering services.
Table 5-1 and table 5-2 report the top 10 industries which hire the largest number
and have the highest percentage of skilled immigrants. In the last 8 years, hospitals,
educational organizations and computer and data processing services have the
largest number of skilled immigrants. Computer and data processing services,
machinery and drugs have the highest percentage of skilled immigrants. In 1980,
1990 and 2000, hospitals and educational organizations hire the largest amount of
skilled immigrants
Next, the industries described above are used to analyze the impact of ACWIA of
1998 on the return of employers and shareholders. The return of each firm of one
certain industry on each trading day and the market return (value-weighted
NYSE/AMEX/NASDAQ) on each trading day are obtained from the CRSP (Center for
Research on Security Prices).
Compared to the result of Lowell and Christian (2000), the list of industries that
will be used to analyze the impact are Computers and related equipment, Computer
and data processing services, Research, development, and testing services, Colleges
and universities, Electrical machinery, equipment, and supplies, Management and
public relations services, Engineering, architectural, and surveying services, Radio, TV,
and computer stores, Security, commodity brokerage, and investment companies.
This list covers the IT and non-IT industries described above.
The event date (day 0 in the analysis) of October 16, 1998 is chosen because the
.
18
bill faced a sudden death in the Congress on October 8 but after a skillful legislative
maneuver by its supporters, the H-lB visa bill made a remarkable, eleventh-hour
comeback as part of the omnibus appropriations bill. Therefore, the news that was
not anticipated by the market would have real effects on the return of those
industries which hire large number and percentage of skilled immigrants. Relatively,
the day that President Clinton signed the bill (October 21) is less important because
everyone has anticipated the information since the bill was passed and the president
would sign it eventually.
30 trading days before and after the event day will be demonstrated to see if the
AWCIA of 1998 has a positive or negative impact on those industries. The procedure
of calculating mean abnormal returns and hypothesis testing are described in section
II.B. The empirical results of all industries are reported in table 6. Individual industry
results are reported in table 7-1 to table 7-10. The summary is the following,
Computers and Related Equipment
- 47 firms in this industry during the analysis.
- Example of the firms are Apple Computer, Sprint, Seagate, Lexmark, Gateway,
Micron Technology, etc.
- The mean abnormal return after the event day (day +1 to day +30) is 16.78%.
Cumulative mean abnormal return is 12.58% during this period.
- Both returns are statistically significant under 1% level in four different tests.
Computer and Data Processing Services
- 581 firms in this industry during the analysis.
- Example of the firms are Oracle, Microsoft, Compaq, Yahoo, Novell, Sandata,
etc.
- The mean abnormal return after the event day (day +1 to day +30) is 18.65%.
.
19
Cumulative mean abnormal return is 19.02% during this period.
- Both returns are statistically significant under 1% level in four different tests.
Research, Development, and Testing Services
- 81 firms in this industry during the analysis.
- Example of the firms are Atlantic Pharmaceuticals, Pacific Biometrics,
Megabios, Opinion Research, Pharmchem, etc.
- The mean abnormal return after the event day (day +1 to day +30) is 15.58%.
Cumulative mean abnormal return is 15.30% during this period.
- Both returns are statistically significant under 1% level in four different tests.
Colleges and Universities
- 5 firms in this industry during the analysis.
- Example of the firms are California Culinary Academy, Computer Learning
- The mean abnormal return after the event day (day +1 to day +30) is -1.83%.
Cumulative mean abnormal return is 22.69% during this period.
- Both returns are statistically significant under 10% level in four different tests.
Electrical Machinery, Equipment, and Supplies
- 346 firms in this industry during the analysis.
- Example of the firms are Texas Instruments, Emerson Electric, Bell Industries,
- Integrated Device Technology, etc.
- The mean abnormal return after the event day (day +1 to day +30) is 22.15%.
Cumulative mean abnormal return is 21.54% during this period.
- Both returns are statistically significant under 1% level in four different tests.
Management and Public Relations Services
- 52 firms in this industry during the analysis.
- Example of the firms are Randers Group, Market Facts, Dental Care Alliance,
.
20
Right Management Consultants, etc.
- The mean abnormal return after the event day (day +1 to day +30) is 11.09%.
Cumulative mean abnormal return is 7.87% during this period.
- Both returns are statistically significant under 1%, 5% and 10% level in four
different tests.
Engineering, Architectural, and Surveying Services
- 24 firms in this industry during the analysis.
- Example of the firms are Cam Designs, Waste Systems International,
Wavetech, etc.
- The mean abnormal return after the event day (day +1 to day +30) is 14.97%.
Cumulative mean abnormal return is 11.45% during this period.
- Both returns are statistically significant under 5% level in four different tests.
Radio, TV, and Computer Stores
- Only 2 firms in this industry during the analysis.
- Example of the firms are Intertan Inc., Grupo Elektra.
- The mean abnormal return after the event day (day +1 to day +30) is 32.13%.
Cumulative mean abnormal return is 27.36% during this period.
- Both returns are statistically significant in four different tests.
Security, Commodity Brokerage, and Investment Companies
- 1361 firms in this industry during the analysis.
- Example of the firms are Smith Barney, Sovereign Bancorp, Fidelity Bancorp,
etc.
- The mean abnormal return after the event day (day +1 to day +30) is 2.37%.
Cumulative mean abnormal return is 3.77% during this period.
- Both returns are statistically significant under 1% level in four different tests.
.
21
It is extremely useful and clear at a glance to draw both returns and compare their
patterns all together. Figure 3 shows the graphs of mean abnormal return (dash line)
and cumulative mean abnormal return (solid line) for those industries. The graphs
clearly show that the returns of those industries are most likely to be affected by the
legislation of ACWIA of 1998. Eight industries in Figure 2 show a similar pattern of
increasing cumulative mean abnormal return after the event day (October 16), the
day which the H-1B visa bill made a remarkable, eleventh-hour comeback). On
October 21, 1998, the day that President Clinton signed the H-1B visa bill did not
show any particular impact which is not surprising because the news has already
been anticipated by the market.
Note that on September 29, 1998 (day -13) shows a significant drop in returns in
all industries. It was because on that day was the Fed cut the interest rate by 25 basis
points for the first time since 1996. However, investors were disappointed about the
Fed’s action because the rate cut was too little and the market was anticipating a
much more decrease of interest rate (see The Wall Street Journal). Beside this, there
was no news that significantly affected the returns of those firms as we can see from
the pattern of the cumulative abnormal return in figure 3.
In sum, the AWCIA of 1998 has a positive impact on the returns of those top H-1B
user industries. Shareholders of high-tech industries (top users of H-1B visa, 80%)
such as Computers and related equipment, Computer and data processing services
enjoy an average 16.78% and 18.65% cumulative excess return in a month after the
Act was passed. On the other hand, shareholders of non-high-tech industries
(remaining balance of H-1B visa) such as Security, commodity brokerage, and
investment companies has an average lower 2.37% cumulative excess return in a
month after the Act was passed.
.
22
Figure 3 Mean Abnormal Return and Cumulative Mean Abnormal Return of Industries:
High percentage and number of skilled immigrants
-.06
-.04
-.02
0
.02
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Security commodity brokerage and investment companies
-.2
0.2
.4
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Radio TV and computer stores
-.1
0.1
.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Engineering architectural and surveying services
-.1
-.05
0
.05
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Management and public relations services
-.1
0.1
.2.3
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Electrical machinery equipment and supplies
-.1
0.1
.2.3
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Colleges and universities
-.05
0
.05
.1.1
5.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Research development and testing services
-.1
-.05
0
.05
.1.1
5
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Computer and data processing services
-.05
0
.05
.1.1
5.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Computers and related equipment
.
23
Robustness
Low-skilled immigrants industry of US
To support the findings, I report US industries which hire low number and low percentage of
skilled immigrants in figure 4 (they can be viewed as non-H-1B users). 13F
14 One can see that the
patterns are different to figure 3. The impact of ACWIA of 1998 which greatly increased the
H-1B visas has no significant effect on those industries which were obviously not benefited
from the increased H-1B visas. For example, these industries are Farm-product raw materials,
Nonmetallic mining and quarrying except fuels, Sawmills, planing mills and millwork, Metal
mining and Bowling centers, etc.
International Factor - Canada, UK and Germany
To check whether the finding is due to international macroeconomic factor, comparison
across countries is necessary. In figure 5, I report the results of mean abnormal return and
cumulative mean abnormal return of high-tech industry in Canada, UK and Germany under
the same time period. The data of Canada is from Canadian Financial Market Research Centre
(CFMRC), UK and Germany are from Global Financial Data. I pick two computer related
industries of US for comparisons.
From figure 5, the result shows that high-tech industries in Canada, UK and Germany do
not have a significant and increasing trend after October 16, the event date. Especially, the
Candadian market is regarded as high-correlated market with the US since both are
commonly viewed as the North American Market. For European markets, the Information
technology sector in UK does not show increasing trend of abnormal return and German
market shows mixed results (not as consistent as US high-tech industry) after the event date.
Hence, the international macroeconomic factor that can affect the finding can be ruled out.
14 Note that some industries hire both large number of skilled and low-skilled immigrants. For example, hospitals, hotels and motels, all construction and eating and drinking places. Hence, for comparison, using the industries that hire large number of medium-skilled or low-skilled immigrants may lead to misinterpret the results.
.
24
Multiple Factor Reuturn Model
This section I choose Fama-Frech (1993) three-factor model as the return-generating process.
The model is:
it i i mt i t i t itR R s SMB h HML (5)
where itR is the rate of return of the common stock of the thi firm on day t ; mtR is the rate of
return of a market index on day t ; tSMB (Small Minus Big) is the average return on the three
small market-capitalization portfolios minus the average return on the three big
market-capitalization portfolios on day t ; tHML is the average return on two high
book-to-market equity portfolios minus the average return on two low book-to-market equity
portfolios; it is a random variable that, by construction, must have an expected value of zero,
and is assumed to be uncorrelated with mtR , uncorrelated with ktR for k j , not
autocorrelated, and homoskedastic. i is a parameter that measures the sensitivity of itR
to the excess return on the market index; is measures the sensitivity of itR to the difference
between small and large capitalization stock returns; and ih measures the sensitivity of itR
to the difference between value and growth stock returns.
Define the abnormal return (i.e., prediction error) for the comom stock of the thi firm on
day t as:
ˆˆˆ ˆ( )it it i i mt i t i tAR R R s SMB h HML (6)
Where the coefficients ˆi , ˆ
i , is and ˆih are ordinary least squares estimates of i , i ,
is and ih .
Figure 6 reports the mean abnormal return and cumulative mean abnormal return of each
industry. From the graphs, the results of the Fama-French three factor model are different
.
25
from the market model of figure 3.
Nonparametric (Semiparametric) Model
Without assuming the functional form of right hand side variables, I estimate the normal
return generating process as:
0( )i i iY G X (7)
where G is some unknow function, ( | ) 0i iE X and 0( | ) ( )i i iE Y X G X . To estimate,
simply:
2
[ ( )]i iMin Y G X
(8)
Hence, the abnormal return can be obtained by getting the conditional expectation
( | )i iE Y X after substrating from iY . In othe words, the abnormal return (prediction error) is
ˆ ( )it itAR Y G (9)
Figure 7 reports the resuts of this model. From the pattern of the graphs, the nonparametric
model seems to catch the impact very well.
Campbell, Lo and Mackinlay (1997) argue that in practice the gains from employing
multifactor models for modeling the normal return of event studies are limited. The reason
for this is that the marginal explanatory power of additional factors beyond the market factor
is small, and hence there is little reduction in the variance of the abnormal return. Hence,
comparing the results of Market model, Nonparametric model and Fama-French three factor
model, my findings confirm their points.
Structural Break Test
To support the findings, I perfrom the breakpoint Chow test to see whether there is a
structural change. The idea of the breakpoint Chow test is to fit the equation separately for
.
26
each subsample and to see whether there are significant differences in the estimated
equations. A significant difference indicates a structural change in the relationship.
Table 9-1 to 9-9 report the results of Chow breakpoint test. On day 0, the F-statistics of
Computers and Related Equipment, Computer and Data Processing Services, Research,
Development, and Testing Services and Electrical Machinery, Equipment, and Supplies
industries are significantly from 0. Hence, the structural break tests support the findings that
there was an impact on the event day.
.
27
Figure 4 Mean Abnormal Return and Cumulative Mean Abnormal Return of Industries:
Low percentage and number of skilled immigrants
-.1
0.1
.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Days
Cumulative Mean Abnormal Return Mean Abnormal Return
Farm-product raw materials
-.1
0.1
.2.3
.4
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Days
Cumulative Mean Abnormal Return Mean Abnormal Return
Nonmetallic mining and quarrying except fuels
-.05
0
.05
.1.1
5
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Days
Cumulative Mean Abnormal Return Mean Abnormal Return
Sawmills planing mills and millwork
-.1
-.05
0
.05
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Days
Cumulative Mean Abnormal Return Mean Abnormal Return
Bowling centers
-.1
0.1
.2.3
.4
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Days
Cumulative Mean Abnormal Return Mean Abnormal Return
Metal mining
-.1
0.1
.2.3
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Hospitals
.
28
Figure 5 High-tech industry in Canada, UK and Germany-.
1-.
05
0
.05
.1.1
5
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Computer and data processing services - US
-.0
5
0
.05
.1.1
5.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Computers and related equipment - US
-.2
-.1
5-.
1-.
05
0
.05
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Telecommunication services - Canada
-.3
-.2
-.1
0.1
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Information Technology - Canada
-.1
-.0
5
0
.05
.1
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Telecommunication services - Germany
-.1
5-.
1-.
05
0
.05
.1
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Information Technology - Germany
-.0
8-.
06
-.0
4-.
02
0
.02
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Telecommunication services - UK
-.4
-.3
-.2
-.1
0.1
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Information Technology - UK
.
29
Figure 6 Fama-French Three Factor Model
-.01
0
.01
.02
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Security commodity brokerage and investment companies
-.2
0.2
.4
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Radio TV and computer stores
-.05
0
.05
.1.1
5
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Engineering architectural and surveying services
-.06
-.04
-.02
0
.02
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Management and public relations services
0
.05
.1.1
5.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Electrical machinery equipment and supplies
-.1
0.1
.2.3
.4
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Colleges and universities
0
.05
.1.1
5.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Research development and testing services
-.02
0
.02
.04
.06
.08
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Computer and data processing services
-.05
0
.05
.1.1
5
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Computers and related equipment
.
30
Figure 7 Semiparametric Model
-.4
-.2
0.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Radio TV and computer stores
-.2
-.1
0.1
.2.3
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Engineering architectural and surveying services
-.2
-.1
0.1
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Management and public relations services
-.2
-.1
0.1
.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Electrical machinery equipment and supplies
-.4
-.2
0.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Colleges and universities
-.2
-.1
0.1
.2
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Research development and testing services
-.3
-.2
-.1
0.1
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Computer and data processing services
-.2
-.1
0.1
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Computers and related equipment
-.01
0
.01
.02
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Day
Cumulative AR AR
Security commodity brokerage and investment companies
.
31
VI. Concluding Remarks
Immigration is a contentious issue in the industrialized nations of the world. Many of
the key issues in the debate on immigration policy are economic. Most attention has
been paid to the potential adverse effect and possible benefits of immigration on
labor market outcomes of employees (Friedberg and Hunt (1995)). Less attention has
been devoted to the impact of immigration on employers and shareholders. This
paper uses the event study analysis to measure the impact of skilled immigration
legislation on the welfare of employers and shareholders. Due to the high demand of
skilled workers (in particular high-tech industries), the American Competitiveness
and Workforce Improvement Act of 1998 greatly raised the available number of H-1B
visas for skilled immigrants. Hence, whether ACWIA of 1998 increases employers'
welfare is of particular interest.
The empirical results show that industries which hire large number or high
percentage of skilled immigrants benefit from the ACWIA of 1998. Employers and
shareholders of those industries enjoy significant positive abnormal returns after the
bill was passed. For example, shareholders of high-tech industries (top users of H-1B
visa, 80%) such as Computers and related equipment, Computer and data processing
services enjoy an average 17.45% and 22.41% excess return in a month after the Act
was passed, respectively. On the other hand, shareholders of non-high-tech
industries (remaining balance of H-1B visa) such as Security, commodity brokerage,
and investment companies and Management and public relations services have an
relatively lower average 5.09% and 10.98% excess return in a month after the Act
was passed.
.
32
References
Bean, Frank D.; Georges Vernez and Charles B. Keely. 1989. Opening and Closing the
Doors: Evaluating Immigration Reform and Control. Washington, D.C.: The Urban
Institute Press and The RAND Corporation.
Becker, Brian E. and Craig A. Olson. 1986. "The Impact of Strikes on Shareholder
Equity." Industrial and Labor Relations Review, 39(3), 425-38.
____. 1989. "Unionization and Shareholder Interests." Industrial and Labor Relations
Review, 42(2), 246-61.
Boehmer, Ekkehart; Jim Musumeci and Annette B. Poulsen. 1991. "Event-Study
Methodology under Conditions of Event-Induced Variance." Journal of Financial
Economics, 30(2), 253-72.
Brown, Stephen J. and Jerold B. Warner. 1980. "Measuring Security Price
Performance." Journal of Financial Economics, 8(3), 205-58.
Note: The symbols $, *, **, and *** denote statistical significance at the 0.10, 0.05, 0.01 and 0.001levels, respectively, using a generic one-tail test.
The symbols (, < or ), > etc. correspond to $,* and show the direction and generic one-tail significance of the generalized sign test.