Changes in Production and Employment Structure and Relative Wages
in Argentina and Uruguay1
Pablo Sanguinetti
Universidad Torcuato Di Tella
Rodrigo Arim
Universidad Torcuato Di Tella and Instituto de Economía,
Universidad de la Republica
Juan Pantano
Fiel and Universidad Torcuato Di Tella
August 2001
Preliminary
1
This paper has been written as a part of a project sponsored by the World Bank called “Patterns of
Integration in the Global Economy: What Does Latin America Trade? What Do Its Workers Do?. We thank
comments received from D. Lederman and two anonymous referees. As usual, all remaining errors are our
exclusive responsibility.
1. Introduction
Since the beginnings of the nineties Argentina and Uruguay have implemented
profound reforms in various areas of economic activity. These reforms, like trade
liberalization and privatization, has changed the structure of production and of
employment and also the relative demand of labor across skill levels. This may explain
the observed rising trend in wage inequality in these countries. In particular, Galiani
(1999) shows that in Argentina, contrary to what has occurred in the OECD countries, it
cannot be asserted that the returns to college graduates have increased during the eighties.
It is only since the beginning of the nineties that there is clear evidence that the college
wage premium have increased. A similar pattern has been observed for Uruguay
(Casacuberta and Vaillant (2001)).
In this paper we combine micro-data, taking from the various household surveys
and macro data, obtained from national accounts, to investigate the effects of different
shocks in aggregate labor demand composition on relative wages. One question we try to
answer is: does trade liberalization play any role in shaping the Argentine and Uruguay
wage structures during the period studied? In particular, extending the analysis presented
in Galiani and Sanguinetti (2000), we test whether those sectors where import penetration
deepened are, ceteris paribus, the sectors where a higher increase in wage inequality has
taken place. We find evidence that supports the hypothesis tested for Argentina but that
rejects it for Uruguay.
Coincidentally with the process of trade liberalization and the relative decline of
manufacturing, in these countries the non-tradable sectors have experienced a significant
expansion in production and employment. This shift in production has had also
significant effects on the relative demand of workers across skill levels. Thus in the
empirical analysis we also study whether changes in labor composition within the nontradable sector have had significant effects on relative wages. The results for Argentina
suggest that relative demand shifts in non-tradable output have had an inequalzing impact
on wage distribution. In the case of Uruguay, results are less clear cut in the sense the
1
observed raised in labor demand in services did not generate substantial differences in
wages across different skill levels.
Finally a large amount of research has sought to evaluate the effect of skilledbiased technological change on wage inequality. As most of the literature (cf. e.g.
Feenstra, 1998) we indirectly identify the magnitude of this determinant through a timing
variant dummy variable for each skill category. We obtain that this factor explains a
significant portion of the increase in wage inequality after we control for individual
characteristics, sectoral dummies, trade liberalization and shock in the composition of
labor demand across nontradables.
The rest of the paper is organized as follows. Section 2 documents the empirical
evidence about wage inequality in Argentina and Uruguay. Section 3 describes the main
change in the structure of production and employment in both countries. In section 4 we
present a simple analytical framework that will guide and motivate the regression
analysis we describe in section 5. Finally, section 6 concludes.
2
2. Trends in wage inequality in Argentina and Uruguay
In this section we study the evolution of the wage structure in Argentina and
Uruguay. In the Argentinian case, the empirical evidence available is from Greater
Buenos Aires, the main urban agglomerate.2 We measure wage differentials by
educational attainment levels. We define the ensuing three skill groups: unskilled (those
individuals who at most have attended high school but have not finished it), semi-skilled
(those that have finished high school) and skilled workers (those that have finished a
tertiary degree)3. Our study excludes self-employees, owner-managers and unpaid
workers because we are only interested in the study of the changes in the wage structure.
The results of the estimation of the wage premia by gender are shown in the figure 1.4
Figure 1: Skilled and semi-skilled workers wage premia
(Base category: unskilled workers)
Source: Galiani (1999).
2
This market covers approximately half of the labor force of the country.
Galiani (1999) shows that this is the relevant classification of educational skills when analyzing the
evolution of wages.
4
These estimations are derived from the coefficients of a wage equation where the dependent variable is
the logarithm of the hourly wages and among the covariates there is a set of educational dummies and a
quadratic function in potential experience. The equations are estimated separately by gender. The
dependent variable is the logarithm of the hourly earnings of the sampled individuals in their main
occupation. For employees, this variable is equivalent to the hourly wages. The schooling group g wage
premium in year t is the expected percentage increase in the wage of a worker whose level of education is g
with respect to the expected wage of an unskilled worker. The yearly data is taken from the October wave
of the Household survey for Greater Buenos Aires (GBA). There are not data tapes available for the years
1983 and 1984.
3
3
For the whole period, the main changes in the wage structure are the following:
the semi-skilled group has become more like the unskilled group as time has passed, that
is, they have seen their wages deteriorated relative to the unskilled group wages.
Additionally, the unskilled group has not seen its wages deteriorate relative to the skilled
workers wages. For example, the male skilled wage premium was 228 percent in 1980,
156 percent in 1991 and 211 percent in 1998 while the male semi-skilled wage premium
was respectively, 87, 44 and 48 percent.
Nevertheless, if the analysis is restricted to the evolution of wages during the
nineties, the period when trade liberalization was deepened, we see a somewhat different
picture. The wages of the semi-skilled group did not deteriorate relative to the unskilled
group wages while both the unskilled and semi-skilled wages deteriorated relative to the
skilled group wages. Indeed, the skilled-unskilled wage premium increased substantially
during the 90s.
Figure 2 illustrates the evolution of the wage premia for the manufacturing
sector.5 Due to sample size considerations we present only an average wage premium by
skill group. It is manifest from the figure that the trends we observe in the manufacturing
sector during the nineties are similar to those we observe for the whole economy. We find
a significant positive trend in the college wage premium. On average, it increased
approximately 7 percentage points per year during the nineties while the secondary
school wage premium slightly decreased but not significantly.6 Thus, overall, we may
conclude that during the nineties, the trends in the wage structure in the manufacturing
sector are quite similar to those for the whole economy.
5
These statistics are derived from the coefficients of a wage equation where the dependent variable is the
logarithm of the hourly wages and among the covariates there is a set of educational dummies, a quadratic
function in potential experience and a gender dummy. The dependent variable is the logarithm of the hourly
earnings of the sampled individuals in their main occupation. The yearly data is taken from the October
wave of the Household survey for Greater Buenos Aires (GBA).
4
Figure 2: Skilled and semi-skilled workers wage premia
in the manufacturing sector, Argentina
(Base category: unskilled workers)
400
350
300
250
200
150
100
50
0
%
Tertiary wage Premium
Secondary school wage Premium
1990 1991 1992 1993 1994 1995 1996 1997 1998
Source: author’s elaboration.
For Uruguay we use the microdata from household surveys for the period 19861997. In this case, we had access to the whole dataset not only the main metropolitan area so
the result extend more naturally to a national (urban) interpretation. Table 1 displays the
logarithmic change in hourly real wages for urban Uruguay for different periods and
different educational levels. We can draw some interesting conclusions regarding the
evolution of relative wages. The overall rate of growth in wages is similar for both periods
(1986-1990 and 1991-1999) and will be our benchmark point of reference to evaluate the
performance of wages for different educational levels. It is interesting to note that in the
second half of the eighties Complete Primary and Complete Secondary were the sectors
with the largest wage increases, while higher educational level wages performed from
moderate growth (Incomplete College) to disappointing 1.1 (complete college) and –0.5
(Professors). The picture is the exactly the opposite in the nineties with higher educational
levels wages growing clearly faster than those corresponding to lower of education like
incomplete/complete primary and secondary.
6
Indeed, like for the entire economy, the rise in the skilled workers wage premium started in 1992. It is
also worth noting that the 1995 value of this statistic is extremely high in the manufacturing sector.
However, it may be even due to sampling variability or mesurement error.
5
Table 1 Logarithmic Change in Real Hourly Wages by
educational level, Uruguay
educational level
1986/1991 1991/1999 1986/1999
Primary school
14.9
9.9
24.8
Incomplete high school
11.4
6.2
17.5
Complete high school
18.1
4
22.1
Technical education
16.7
8.7
25.4
Teachers
-0.5
25.8
25.3
Incomplete collage
9.2
18.3
27.5
Complete collage
1.1
32
33.1
Overall
11.7
11.5
23.2
Source: Arim and Zoppolo (2000)
As we did with the case of Argentina, we will aggregate the different levels of
education into three groups. Still as in Uruguay incomplete college wages are closer to
complete college ones, we prefer to categorize Uruguayan skill groups differently. So, the
high skill group for Uruguay includes complete as well as incomplete college workers.
Workers with complete and incomplete high school compose the semi skill group. The
remaining workers (those with at most complete primary education) are in the low-skill
group.
Figure 3 shows the evolution of the wage premia at the manufacturing sector for
Uruguay. We present the evolution of the collage and high school premium with respect
to the wage of a worker with only primary complete school. We observe that also in
Uruguay, for both men and women, there has been an increase in college wage premia
6
especially in the nineties. Thus, the observed pattern for both countries is very similar
also at the manufacturing sector.
7
. Figure 3: Skilled and semi-skilled workers wage premia
in the manufacturing sector, Uruguay
(Base category: unskilled workers)
A-Men
College premium
High School premium
350%
300%
250%
200%
150%
100%
50%
0%
1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
B- Women
(Base category: unskilled workers)
College premium
High School premium
300%
250%
200%
150%
100%
50%
0%
1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
Source: authors elaboration based on household surveys
8
3. The changes in the production and employment patterns
In this section we want to revise the evidence regarding changes in sectoral composition
of production and employment in these economies, which could potentially explain part
of the described movements of wage inequality. We start analyzing the evolution of
production structure for both countries. Afterwards, we analyze the change in the
employment structure and the differences in the sectoral requirements of human capital.
The type of question we want to address here are: which are the sectors where
production and employment expanded and which were the declining sectors? How does
the allocation of labor skills varied across time and across sectors of the economy?
3.1. Production patterns
A. Argentina
The evolution of aggregate GDP for Argentina shows a marked contrast between the
eighties and the nineties. During the first decade GDP falls 10% while in the nineties
GDP raises approximately 37% (see figure 4). This result is not a surprise taking into
account the deep and far reaching reforms that the country has experienced since 1991.
Within these reforms inflation stabilization has had a strong impact in spurring aggregate
demand, which has plunged since late 1988 as a consequence of the hyperinflation
process suffered by the economy. Nevertheless, the strong growth that we observe since
1991 was not only a cyclical recuperation from a low level of output. Several estimations
(see Kydland and Zarazaga, 1997) suggest that the potential GDP of Argentina have
change its tendency during this period as a consequence of the increase in capital
accumulation that was spurred by the reforms policies such as trade and investment
liberalization.
9
Figure 4
GDP (pesos 1993)
GDP
300,000,000
250,000,000
200,000,000
150,000,000
100,000,000
50,000,000
Year
0
Source: INDEC.
Beyond what has happen with aggregate GDP, from the point of view of this paper, we
are interested in analyzing the changes that may have occurred in the structure of
production during these two periods. Table 2 presents the structure of GDP disaggregate
at 1 digit of the ISIC since 1980.
10
Table 2. GDP Structure. Argentina
in %
Year
Transport
Agricult
Electricity,
Retailing
Financial Public
and
ure and Mining Industry Water and Construction and Hotel
and real Administ
telecomuni
Fishing
Gas
services
state
ration
cation
1980
5.12
1981
5.51
1982
5.98
1983
5.90
1984
5.82
1985
6.11
1986
5.71
1987
5.42
1988
5.96
1989
5.86
1990
6.53
1991
6.16
1992
5.67
1993
5.49
1994
5.56
1995
6.03
1996
5.65
1997
5.26
1998
5.47
1999
5.72
2000
5.61
Source: INDEC.
1.42
1.46
1.46
1.45
1.41
1.46
1.27
1.33
1.43
1.53
1.61
1.50
1.53
1.59
1.71
2.04
2.02
1.89
1.74
1.70
1.88
21.10
19.39
19.39
20.20
20.43
19.71
20.50
20.18
19.65
19.47
19.25
19.24
19.72
19.50
19.18
18.29
18.48
18.68
18.20
17.28
16.90
1.46
1.54
1.66
1.72
1.83
2.00
1.93
1.98
1.88
1.92
2.12
1.98
1.98
2.08
2.17
2.39
2.36
2.37
2.44
2.60
2.79
8.52
7.80
7.25
6.93
6.07
5.53
6.20
6.92
6.84
5.54
4.52
5.35
5.78
6.05
6.03
5.44
5.60
6.04
6.28
5.93
5.29
19.10
18.18
17.17
17.45
18.22
17.37
17.28
17.00
16.65
16.64
17.05
17.84
18.17
17.76
17.85
16.96
17.36
17.83
17.64
16.97
16.70
5.82
5.72
6.01
6.03
6.34
6.64
6.69
6.74
6.75
7.16
7.12
7.11
7.33
7.29
7.57
7.91
8.03
8.27
8.61
8.75
8.90
18.21
20.20
20.18
19.55
19.00
19.67
19.59
19.35
19.38
19.12
18.59
19.14
18.90
19.63
20.24
20.59
20.74
20.61
21.11
21.79
22.27
19.26
20.20
20.91
20.76
20.89
21.52
20.83
21.07
21.46
22.77
23.22
21.68
20.91
20.59
19.70
20.35
19.76
19.04
18.49
19.26
19.66
Two trends are clearly observed from the data. On one hand, industry has been losing
importance in terms of its participation in GDP. It was around 20% in the beginning of
the 80s and ended up with a share near 17 % in year 1999- 2000. Most of the decline in
this participation has occurred during the nineties; in 1991-1992 industry participation
was around 19.5, which was pretty close to the average of the eighties. The second clear
trend was the increase in participation in the service sector. Its share was 43% of the GDP
in 1980 and increased to 48% in year 2000. Again, most of the increase in this share has
happened during the nineties. Within the service category the ones that increase the most
were electricity, gas and water, transport and telecommunication and financial services
11
total
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
and services to firms. Clearly the raise in the first two activities is associated with the
privatization policies adopted by Argentina at the beginning of the nineties; on the other
hand, the expansion of financial services has been a consequence of the stabilization
policies which implied an important increase in the degree of monetization of the
economy.
Regarding the other sectors, primary, the least important of all, experienced an erratic
behavior in terms of its share, reflecting the volatility of prices of these commodities. For
example, Agriculture and Fishing’s share remained pretty stable between the two extreme
years of the period while we observe temporary rises in some years in concordance with
the behavior of agricultural prices. On the other hand, we observe an increasing trend in
mining, which took place since the beginning of the nineties, though it was partially
reverted at the end of the decade when international prices for these products suffered
sharp declines.
Finally, construction shows a declining tendency over the period, which was mainly
produced during the eighties. The share corresponding to public sector activities has also
remained stable between the extreme years of the period though it rose at the end of the
eighties as a consequence of the sharp recession that affected the rest of the activities
during the hyperinflation episode.
B. Uruguay
The evolution of GDP in Uruguay during the last 15 years shows a gradual process of
recuperation of the economy after the stagnation suffered in the first half of the eighties.
Thus between 1986 and 1991 the economy grew at an annual rate of 1.8%. Afterwards in
the nineties the growth rate accelerated, reaching an average value of 4.1% between 1991
and 1999.
12
The structure of production has also undergone significant changes in last two decades
as illustrated by the data presented in Table 3, where we show a disaggregation of GDP
by 1-digit sector of the ISIC classification for selected years.
Table 3: GDP Structure. Uruguay
1986
1988
1994
1999
12.7
8.7
7.7
5.5
0.1
0.1
0.2
0.2
29.7
26.5
18.3
16.0
4 Electricity, gas and water
3.6
2.6
3.1
3.8
5 Construction
2.7
3.6
5.7
5.8
12.6
14.6
16.5
13.5
6.4
6.5
6.9
8.4
8 Financial Institutions
18.3
21.1
22.8
26.2
9 Public sector and other services
14.0
16.2
18.8
20.6
1 Agriculture/cattle
2 Mining
3 Manufacturing
6 Retailing ad hotel services
7 Transportation and telecommunications
Source: Banco Central del Uruguay
As was the case in Argentina, manufacturing is the sector for which we observe the
most significant fall in GDP participation between 1986 and 1999. Industry represented a
29.7% of GDP in 1986 and fell to around 17.0% in 1999. Again, similarly with
Argentina, most of the fall in the participation took place in the nineties, especially in the
first half of the decade. The other sector that losses participation is primary production
mainly by the reduction in participation of Agriculture and cattle (fishing is negligible in
the case of Uruguay). The share of primary production was 12.7% in 1986 and went
down to 5.5% in 1999. Compared to industry, the decline of the share in this sector has
been a more continuous process, which took place along the whole period.
The sectors where we find an expansion in production above the average are those
related with certain services. This is notably the case of financial institutions and services
to enterprises, for which the share went up from 18.3% in 1986 to 26.2 in 1999.
13
Construction was also another sector that increased its participation in total production,
especially during the nineties. The activity associated with public sector and other
services (sector 9) has also expanded its production above the average, increasing its
share from 14.4% in 1986 to 21.3% in 1999. On the other hand for retail, restaurants and
hotel services we find similar shares in 1986 and 1999, though there was a temporary
raise in it at the end of the eighties and beginning of the nineties.
3.2. The evolution of employment structure
To what extend does the above changes in the production structure has been translate to
the structure of employment? In this section we will look at this issue describing the
evidence on the sectoral allocation of labor across major industry and services sectors of
the economy. The aim is to identify shift in labor demand across sectors of the economy
that may have been induced by the above-described change in the structure of production.
A.
Argentina
The evidence from the Permanent Household Surveys shows that there was a
significant decline in the employment share for almost all manufacturing sectors during
the period under analysis. For the aggregate of industry the reduction was equal to 10
percentage points (see Table 4) and it occurred mostly during the nineties. This fall was
compensated by increases in some services, mainly business and financial services,
which expanded overall from 7.8% in 1985 to 11.5% in 1999. Table 5 shows that the
reduction in manufacturing employment is more important for Textile and Footwear,
falling from 8.2% in 1985 to 3.5% in 1999. These sectors are the usual reference as an
example of the negative impact of trade liberalization on employment in the industrial
countries.
Given that the survey coverage is only urban, primary sector employment share is
substantially underestimated and, as a consequence, manufacturing and services
employment shares are overestimated. Figure 5 shows the employment level evolution
14
for the 8 aggregates highlighted in the previous table. As we can see, not only the
manufacturing sector lost share in total employment but there was also a significant
decline in the absolute level of employment.
Table 4 Employment share by selected sectors , Argentina
Sector
Primary products
Manufacturing Sector
Food Drinks and Tobacco
Textile and Footwear
Chemical Productos
Metalic products
Other Industries
Electricity, Gas and Water
Construction
Trade, Hotels and Restaurants
Major Trade
Retail trade
Hotels and Restaurants
Transportation and Communications
Transportation
Transportation related services and comunications
Bussinnes and Financial Services
Finance
Real estate and businnes services
Social and Personal Services
Public Administration and Defense
Teaching
Social and Health services
Other social services
Repair services
Housekeeping
Other personal services
Source: Authors calculations based on EPH
15
1985
1990
1994
1999
0.3
26.1
3.5
8.1
3.1
6.3
5.1
0.2
6.9
18.6
3.6
13.1
1.9
7.2
5.8
1.4
7.8
1.6
6.2
32.9
4.0
5.2
4.0
3.8
3.1
8.2
4.6
0.4
24.0
3.3
6.2
3.1
6.1
5.2
0.8
6.2
20.1
3.7
13.9
2.6
8.0
6.4
1.6
7.2
2.3
4.9
33.2
5.1
5.7
5.1
4.0
2.5
9.2
1.6
0.4
21.3
3.5
4.1
2.5
6.3
4.9
0.7
7.1
20.7
5.5
12.1
3.1
9.2
6.6
2.6
9.2
2.8
6.5
31.5
4.7
6.3
4.9
3.7
3.1
7.6
1.2
0.4
17.3
2.8
3.4
2.9
4.6
3.5
0.4
7.6
19.3
4.6
11.5
3.3
9.7
6.6
3.1
11.5
2.8
8.7
33.8
5.1
7.0
5.4
4.4
2.8
7.8
1.4
Figure 5
Evolution of Employment by selected sectors (thousands), Argentina
('000)
1200
2500
1100
2300
Social and Personal Services
Manufacturing sector
1000
2100
900
1900
1986
80
1988
1990
1992
1994
1996
1998
Electricity, Gas
and Water
1986
350
70
300
60
250
50
1990
1992
1994
1996
1998
1990
1992
1994
1996
1998
1992
1994
1996
1998
1992
1994
1996
1998
Construction
200
1986
1988
1990
1992
1994
1996
1998
950
850
1988
1986
1988
500
Trade, Hotels and
Restaurants
450
Transport and Communications
400
750
350
650
1986
500
1988
1990
1992
1994
1996
1998
Bussines and Financial Services
1986
50
400
1988
1990
Primary Sector
40
year
300
1986
30
1988
1990
1992
1994
1996
1998
1986
1988
1990
Source: Galiani and Sanguinetti (2000)
B.
Uruguay
Similarly to Argentina, in Uruguay manufacturing is the most affected sector in
terms of lost of employment (see Table 5). Again for this economy the bulk of the decline
is concentrated in the nineties. As we see from the table, between 1986 and 1990
manufacturing sector employment share remains relatively stable or even increases
slightly. The 5-percentage points decline in the share occurred between 1990 and 1999,
which seems moderate compared to the case of Argentine.
16
Table 5 Employment share by sectors, Uruguay
Sectors
1986 1990 1994 1999
Agriculture, Hunting, Fishing
Mining
Manufacturing
Electricity, Gas and Water
Construction
Retail Trade, Restaurants and Hotels
Transportation, Storage and Communication
Financial and Bussinnes Services
Personal, Social and Comunal Services
4.3
0.2
20.6
1.7
5.1
17.8
7
4.5
38.8
3.4
0.2
21.1
1.5
6.6
17.8
6
4.8
38.5
4.4
0.1
19.2
1.2
7.3
19.2
6
5.9
36.6
3.9
0.1
15.9
1
8.4
19.8
6.1
6.5
38.3
Source: authors elaboration based on ECH and information from BCU
The most dynamic sectors displaying increases in their employment shares are
Construction and Financial and Business services. However, preliminary and more
disaggregated computations show that the majority of this increment is due to business
services and not to financial services. The other sectors remain relatively constant all
along the period.
3.3 Human capital requirements across sectors
In this section we briefly summarize the changes in human capital requirements across
different sectors for Argentina and Uruguay. We will use the same three educational
levels defined for each country in section 2.
A. Argentina
Table 6 displays the human capital requirements by sector corresponding to Argentina.
We observe substantial movements in all activities towards a more high skilled intensive
form of production. This is reflected in a widespread decline in the low education
intensity of employment across all sectors7. This process coincides with the increase in
relative supply of skilled workers that took place during this period (Galiani (1999)). On
7
We neglect Primary Products given the urban coverage of the survey and other sectors with a very small
number of observations like Electricity, Gas and Water.
17
the other hand, we see that sectors that ranked among those with higher human capital
requirements in 1986, like Business and Financial Services or Personal and Social
Services, still remain in that condition in 1999. On the other extreme, manufacturing was
both, in 1986 and in 1999, one of the activities that most intensively used low skilled
workers.
This different human capital requirements across sectors could have had
important consequences on the behavior of relative demand for skilled and unskilled
labor (and for relative wages), if, as we observed in Argentina, manufacturing
employment falls while the employment of Business and Financial Services rises.
18
Table 6 Human Capital requirements by sectors , Argentina
Education level
1986
Sector
Primary products
Manufacturing Sector
Food Drinks and Tobacco
Textile and Footwear
Chemical Productos
Metalic products
Other Industries
Electricity, Gas and Water
Construction
Trade, Hotels and Restaurants
Major Trade
Retail trade
Hotels and Restaurants
Transportation and Communications
Transportation
Transportation related services and com.
Bussinnes and Financial Services
Finance
Real estate and businnes services
Social and Personal Services
Public Administration and Defense
Teaching
Social and Health services
Other social services
Repair services
Housekeeping
Other personal services
1999
Low
Medium
High
Low
Medium
High
56%
76%
84%
82%
63%
73%
74%
38%
87%
73%
65%
74%
88%
73%
77%
57%
29%
33%
27%
64%
54%
17%
41%
63%
84%
97%
73%
25%
21%
15%
17%
29%
22%
22%
62%
9%
23%
30%
24%
12%
23%
19%
36%
48%
57%
43%
21%
33%
44%
21%
24%
15%
3%
22%
19%
4%
1%
2%
8%
5%
5%
0%
4%
3%
4%
3%
1%
4%
3%
7%
24%
10%
30%
15%
13%
39%
38%
13%
1%
0%
5%
59%
60%
71%
68%
49%
50%
65%
55%
83%
58%
43%
61%
67%
61%
69%
44%
20%
11%
23%
48%
39%
13%
28%
55%
69%
86%
52%
6%
33%
26%
30%
39%
40%
30%
30%
14%
36%
45%
34%
30%
33%
27%
46%
50%
65%
45%
28%
42%
33%
22%
35%
27%
12%
43%
35%
7%
3%
3%
12%
10%
5%
15%
3%
6%
13%
5%
3%
6%
4%
10%
30%
24%
32%
24%
18%
54%
50%
10%
4%
2%
5%
Source: Authors calculations based on EPH
B. Uruguay
The change in human capital requirements in Uruguay is very similar to that of
Argentina. On the supply side, there is a significant rise of the educational level of the
labor force. Coinciding with this phenomenon, almost all sectors increase their tertiary
labor intensity (see Table 7). We also observe that sectors that were more intensive in
tertiary level employment in 1986, mainly services sectors, remained also very intensive
19
in the use of this factor in 1999 (e.g. Real Estate and Business services, Finance and
Insurance and Social services, etc). In addition these sectors were those where
employment expanded the most. We may suspect that this change in relative demand
could have affected relative wages between skilled and unskilled workers. We
empirically investigate this hypothesis in section 4.
Table 7 Human Capital requirements by sector, Uruguay
Education Level
Sectors
1986
1999
Primary Secondary Tertiary
Primary Secondary Tertiary
Agriculture, hunting, mining
77.6
Food Drinks and Tobacco
57
Textile and Leather
49.6
Wood
46.3
Paper and paper products and printing35.2
Chemical products
43.2
Metalic products, machines and eq. 38.5
Other manufacturing iindustries
38.1
Eletricity, Gas and Steam
43.1
Water and Hidraulic Projects
40
Construction
69.7
Major Trade
40.5
Retail Trade
40.8
Restaurants and Hotels
58.2
Transportation and Storage
49
Comunications
33
Finance and Insurance
13.2
Real Estate and bussiness services 14.9
Public Administration and Defense 42.5
Social Services and other Comun. Serv.
23.4
Entertaining
43.4
Personal Services
67.3
20
40.5
47.6
52.7
57
51.8
58
58.6
50.1
49.9
28.8
52.9
55.2
40.2
48.8
57
74.5
66.8
49
40.6
50
31.7
Source: authors calculations based on ECH microdata
20
2.3
2.5
2.8
0.9
7.8
6.7
4.5
3.3
6.9
10
1.5
6.7
3.9
1.6
2.3
9.9
12.3
18.3
8.5
36
6.6
0.9
59.4
35.3
31.8
29.3
17
23.7
26.4
17
18.7
36.8
54.6
18.6
19.7
30.6
31.2
11.7
3.5
10.2
27.2
12.7
25.5
49.6
34.5
59.7
65.4
68.3
69
60.4
67.5
75.7
60
49.6
43
67.7
70.8
62
61.6
66.4
57.2
54.9
57.9
39.8
54.3
48.3
6.1
4.9
2.8
2.5
14.1
15.8
6.1
7.3
21.4
13.6
2.4
13.6
9.5
7.4
7.2
21.9
39.3
35
14.9
47.5
20.2
2.1
4. Change in the production and employment structure and wage inequality: a
theoretical background.
We consider the simplest labor market model where there are two labor inputs
represented by skilled and unskilled workers. Figure 6 shows the determination of
relative employment and relative wages between the two kinds of workers at a
competitive equilibrium, after a supply (panel a) and demand (panel b) shock.
Figure 6
S’
S
S
W1
h
W
1W
W0
0
D’
D
D
E’ E
E
Panel a
E’
Panel b
As shown in panel a, an exogenous increase in the supply of skilled workers will reduced
the relative wage of those workers. On the other hand, if the demand curve moves
upwards, the relative wage and the participation in the employment of skilled workers
rises (panel b of figure 6).8 Since the evidence points out that in Argentina and Uruguay
the relative supply shifted in the nineties in favor of skilled workers while, at the same
time, the wage differential rose, the change in the composition of the labor supply cannot
be a plausible explanation behind the observed change in the wage inequality in both
countries. To explain the observed facts we have to bring in stories associated with
changes in the relative demand. Indeed a focal point in the literature about the evolution
of wages differentials in the last two decades has been to identify the different forces that
21
potentially explain the change in the structure of labor demand in favor of skilled
workers.
Several arguments can be distinguished. Some authors have stressed the consequences
of trade liberalization on labor markets (Lawrence and Slaughter (1993), Leamer (1995),
(1998)). The effects of trade on wage inequality are analyzed in the simple framework of
Hecksher-Ohlin (H-O) model with two tradable goods and two non-tradable labor factors,
skilled and unskilled work. In this context, the Stolper-Samuelson theorem, expresses the
basic transmission mechanism that relates the international trade and the wage structure
of a country. Its simplest version establishes that a fall in the relative price of one of the
goods leads to a reduction in the real returns of factor that is used intensively in its
production and a concomitant increase in the returns of other factor. Thus, if international
competition causes a reduction of prices of products that used intensively low-skilled
labor, then we would observe a raise in the wage skilled premium. The (H-O) model
assumes perfect mobility of workers across sectors so it is not a good framework to look
at industry sectors within a given country. Galiani and Sanguinetti (2001) show that a
negative effect of trade liberalization on low skilled relative wages can also be expected
in a economy where it is assumed that low skilled labor is less mobile compared to high
skilled labor. This allows to empirically identifying the effect of trade shocks on the
structure of wages using cross industry data.
Beyond trade liberalization, other authors have argued that the changes in the relative
demand for labor are associated to alterations in the overall productive structure of the
economies that evolves towards sectors that use skilled labor more intensively (Bernard
and Jensen 1998). In special, the lost of participation of manufacturing sector in the GDP,
and the concomitant raise of certain type of skilled–intensive services would be a process
that generates a reduction in the relative demand of less skilled workers.
8
Of course, the size of the change in the relative wage and employment will depend on the elasticity of
supply and demand curves.
22
Finally, another key explanation articulates around the notion of biased technological
change, highly complement of skilled labor, but substitute of unskilled labor (Acemoglu,
2000; Berman and Bradford (1998), Berman et al (1998)). This type of shock would be
the main force shifting the relative demand curve upwards to the right, increasing the
wage premium of skill workers. These authors argued that in the last two decades there
has been an acceleration of technological change that determined an increase of the
marginal productivity of skilled workers greater than the registered by unskilled workers.
This phenomenon would explain the increase of the wage differentials in favor of more
skilled worked
5.
An empirical test of the impact of trade liberalization and changes in non
tradable production structure on wage inequality.
In this section we will try to investigate empirically the determinants of relative wages in
Argentina and Uruguay. In particular we will try to assess whether some the forces we
identify in the previous section have any bearing with the behavior of actual data.
To do this, we start to focalize on the effect of trade liberalization on the structure of
labor demand and relative wages in the manufacturing sector. Afterwards, we generalize
our analysis incorporating the non-tradable activities so as to take into account the effect
of changes of production and employment patterns in the whole economy on wage
inequality. Finally, from the estimation of the time varying dummies for the skill levels
we assess the impact on wage inequality of skilled bias technological change.
5.1 Relative wages and trade liberalization
In this section we study whether the deepening of trade liberalization has had any
identifiable impact on the structure of wages in Argentina and Uruguay. Specifically, we
test, using micro data, whether or not those manufacturing sectors where import
penetration deepened are, ceteris paribus, the sectors where occurred a higher increase in
wage inequality by skill group. Galiani and Sanguinetti (2000) show that the degree of
import penetration has increased in most manufacturing sectors in Argentina during the
23
nineties. This rise in foreign competition was not uniform across sectors. Thus, we are
able to investigate whether, after we control for several individual characteristics, it is the
case that relative wages widened comparatively more in those sectors that faced strongest
competition from foreign markets.
In order to test the hypothesis that import penetration plays a role in shaping wage
inequality, we estimate the coefficients of the following regression function:
Log ( wijt )   dsijgt  gt   dsijgt m jt  gm   dtijct ct  f t (ageijt ) dsexijt  t  ct   j  uijt
g _1
g
(1)
c _1
where dsijgt is a dummy variable that indicates schooling group g in period t, and gt is a
schooling effect in period t; mjt is the logarithm of the ratio of imports to gross value
added in the manufacturing sector j in period t. dtijct is a dummy variable that indicates
tenure group and ct is the tenure effect in period t. The tenure groups are: (0,1), [1,5),
[5,10), [10,20) and [20,20+). ft(ageit) is a non-linear function of the age of individual i in
period t, which is linear in the coefficients to be estimated. dsexijt is a dummy variable
indicating the gender of individual i and t is the gender impact on wages in period t; ct is
the intercept in period t (the period effect); j is the sector fixed-effect, and uijt is the
error term for individual i working in sector j during period t.
The dependent variable is the logarithm of the hourly earnings of the sampled
individuals in their main occupations. The schooling groups are the unskilled group, the
semi-skilled group and the skilled group defined in section 2 (note that skill
categorization differs between Argentina and Uruguay). For Argentina, the micro data on
wages comes from the household survey for the period 1992-1999 for both waves of the
year (May and October). For Uruguay we use the Continuous Households Survey (ECH)
(conducted annually, each one all along the year) corresponding to the 1987-1997
period.. Thus, in each case, the period effect refers to the wave-year or year effect. The
Argentine data on imports, exports and value added by two-digit sector is taken from the
Argentine International Trade Commission. In the case of Uruguay we have taken trade
24
data from Data Intal and value added data from Annual Manufacturing Survey from INE
(Instituto Nacional de Estadística, Uruguay) .
We estimate equation (1) by sampling only the workers of the manufacturing
sector because they are the only group of workers for which the measure of import
penetration adopted presents variability. For Uruguay we previously transform trade data
from ISIC Rev 3 to ISIC Rev 2 in order to compute import penetration ratios with
production data that it is only available at ISIC Rev 2 disagreggation. For Argentina we
work with the 2-digit sectors defined by ISIC Rev 3.
Thus, under the specification adopted for our test, the schooling group g wage
premium in sector j in year t is given by WPjgt = 100 [Exponential(gt + (gm - bm) mjt) –
1], where bm is the estimated coefficient in the regression function 1 for the educational
base category. Consequently, at a first stage, the set of gm are the parameters of interest
in our study. Given our hypothesis, that is, that the relative wages widened comparatively
more in those activities that faced strongest competition from foreign markets and the
evidence gathered in section 2, we expect the difference among the coefficients of the
skilled group and the other two skill groups to be positive. Additionally, we may also
expect these two differences to be statistically similar.
Note that our estimate of the impact of import penetration on wage inequality are not
necessarily an estimate of the whole effect of the former on the latter, that is, it is not
necessarily an estimate of the general equilibrium effect which may not be identifiable.
For example, if trade liberalization shifts labor demand against the unskilled in some
manufacturing sectors and labor is highly mobile, it would be the case that the wages of
the unskilled workers are adjusted in every sector of the economy and hence, the
correlation between the degree of import penetration and wage differentials by sector
vanishes. However, as shown in Galiani and Sanguinetti (2000), under certain
technological conditions or rigidities in the adjustment of the economy, an increase in
import penetration may widen income inequality relatively to the rest of the economy in
the sectors affected. Our test evaluates the existence of these differential effects in the
manufacturing sectors. If we do not find any effect, it is still plausibly, at least
25
theoretically, that import penetration may be shaping wage inequality. Instead, if we do
find an effect from the degree of import penetration on wage inequality, this effect may
not necessarily be an estimate of the general equilibrium effect: it would just be the
identifiable effect.
Note the similitude of our regression model and the wage curve model of
Blanchflower and Oswald (1994). We control both for period fixed-effect and sectorfixed effect. Thus, our model does not provide information about the level of wages by
sector because we are conditioning our estimates on the sample means by sector. In our
model the curve would be drawn in the plane of wage premium and sector import
penetration instead.
It is worth noting that in the specification of the regression function (1) we control
for any aggregate shock that affect wages homogeneously. Thus, for example, if inflation
affects all wages in the same way, it would be captured by the period effect. If instead we
have that some other determinants, for example, technological change, that affects wages
differently by skill group, it would be captured by the wage premium that we allow
varying by period. The latter is an important feature of the specification adopted, which
permits us to estimate the significance of these residual factors affecting relative wages.
On the other hand, the set of parameters gm should only capture the impact on wages of
the sector import penetration.
Galiani and Sanguinetti (2000) show that the estimated coefficients for the variables
controlling for individual characteristics (education level, age and tenure) are as
expected. Next, we report the main results from Galiani and Sanguinetti (2000) for the
Argentine case and new results based on the extension of their methodology to Uruguay.
Table 8 and 9 (for Argentina an Uruguay, respectively) presents the estimated
coefficients of the parameters of interest, those associated with the interactive variables
of education levels and import penetration. The reported standard errors are consistent
standard errors although the errors in the regression function (1) may lack independence.
26
In particular, they are robust to the problem of random group or cluster effects in the data
(cf. e.g. Huber, 1967 and Moulton, 1986).
In the case of Argentina, the coefficients of the interactive import penetration
variable corresponding to the three education level are positive and significant, Galiani
and Sanguinetti (2000) show that this result is robust to alternative specifications (like
controlling for export penetration and for changes in sector prices). Most important, the
coefficient of the skilled group is positive and higher than the coefficient of the other two
skill groups, which have similar estimated values. Thus we find evidence that shows that
in those manufacturing sectors where the import penetration increased the most, wage
inequality also widened relatively more in favor of the most skilled workers.
Table 8: Coefficients (standard errors) of import penetration-skill interaction
variables on wages by skill group, Argentina.
Variable
Coefficient
Robust
standard
error
Unskilled
dummy
*
0.067
0.035 **
Semi-skilled dummy *
0.060
0.035 *
0.125
0.048 ***
import penetration
import penetration
Skilled dummy * import
penetration
Notes: *** if the coefficient is statistically different from zero
at the one percent significance level. ** if the coefficient
is statistically different from zero at the five percent significance
level. * if the coefficient is statistically different from zero at the
ten percent significance level.
27
Table 9: Coefficients (standard errors) of import penetration-skill interaction
variables on wages by skill group, Uruguay.
Variable
Coefficient
Robust
standard
error
Unskilled
dummy
*
import penetration
Semi-skilled dummy *
import penetration
Skilled dummy * import
.001172
.007361
5
.017411
.00706*
2
*
.008952
.013175
penetration
Notes: *** if the coefficient is statistically different from zero
at the one percent significance level. ** if the coefficient
is statistically different from zero at the five percent significance
level. * if the coefficient is statistically different from zero at the
ten percent significance level.
The identified effects for Uruguay are rather weak. Particularly, only the semi-skilled
import penetration interaction is significant and the implied effect on wage inequality is
not so clear for this case. If anything, import penetration will be acting against the trend
observed in section 2 where high skilled wages raised relative to semi-skilled wages.
Still this unintuitive result disappears when, as we will see next, we incorporate the
nontradable sectors in the empirical analysis. Overall, the weak results regarding the
import penetration variable in Uruguay (as compared to Argentina) is also consequence
of the fact that data from the Continuous Household Survey allows disaggregate the
manufacturing sector in only nine activities (we have twenty one in the case of
Argentina).
We partially conclude that there is scope for trade liberalization to explain the
increase of the skilled group wage premium during the 90s only for Argentina. Thus, at
least partially, the aggregate trends on wage differentials we presented in section 2 may
be explained by the impact on trade liberalization on wages. However, the identified
28
effect of trade liberalization on wage inequality does not explain much. Even though the
average (weighted by employment) imports to sector value added increased
approximately 80 percent during the period studied, the average identifiable increase in
the skilled wage premium due to trade liberalization in the manufacturing sector is
approximately 5 percentage points, which is only 10 percent of the increase in the skilled
wage premium during the same period.
5.2 An empirical extension including the non tradable sectors.
We now turn to the extension incorporating non-tradable sectors. The basic idea is to extend
the estimation strategy to allow the inclusion of workers from non-tradables sectors in order to
further investigate the robustness of previous results. Thus, we will try to detect additional
effects on wage inequality that will be operating through shocks channeled by changes in nontradable labor demand. In order to do so, we extend equation (1) in the following way,
Log ( wijt )   dsijgt  gt   dsijgt m jt  gm   dt ijct  ct  f t (ageijt ) dsex ijt  t  ct   j   dsijgt nkt  gk  u ijt
g _1
g
c _1
g
k
To capture effects from changes in relative labor demand in non-tradable sectors we have
added KG interactions between the logarithmn of each nontradable sector share in total
GDP nkt and the three skill level dummies. The kt ‘s will give us these estimated effects.
Table 10 displays the results for Argentina.
29
(2)
Table 10: Coefficients (standard errors) of import penetration-skill and nontradable sectors share in total GDP-skill interaction on wages by skill group,
Argentina.
Variable
Coefficient
Standard
Error
0.0082
-0.0005
0.0608
0.0159
0.0190
0.0266
-0.0345
-0.0252
0.0179
-0.0069
-0.0016
-0.0425
0.0163
0.0192
0.0308
0.0213
0.0347
0.0330
**
-0.0359
0.0076
0.0465
0.0185
-0.0968
-0.0605
0.0203
0.0238
0.0335
0.0237
0.0373
0.0358
*
0.0269
0.0113
0.1826
0.0602
0.0725
0.2107
0.0277
0.0324
0.0418
0.0296
0.0438
0.0413
Import Penetration interactions
with low-skill
with semi-skill
with high-skill
**
Non Tradable sectors interactions
with low-skill
Electricity, Gas and Water
Construction
Wholesale & Retail Trade, Restaurants and Hotels
Transports and Comunications
Financial and Real Estate Intermediation
Other personal and social services
with semi-skill
Electricity, Gas and Water
Construction
Wholesale & Retail Trade, Restaurants and Hotels
Transports and Comunications
Financial and Real Estate Intermediation
Other personal and social services
***
*
with high-skill
Electricity, Gas and Water
Construction
Wholesale & Retail Trade, Restaurants and Hotels
Transports and Comunications
Financial and Real Estate Intermediation
Other personal and social services
***
**
*
***
Notes: *** if the coefficient is statistically different from zero at the one percent significance level. ** if the
coefficientis statistically different from zero at the five percent significance level. * if the coefficient is
statistically different from zero at the ten percent significance level.
We can see that after controlling by non-tradable sector relative demand shifts, the
coefficients on import penetration interactions are reduced. In fact only the high skillimport penetration interaction remains positive and significant.
30
Table 11: Coefficients (standard errors) of import penetration-skill and nontradable sectors share in total GDP-skill interaction on wages by skill group,
Uruguay.
Variable
Coefficient
Standard
Error
-0.0146
0.0037
-0.0033
0.0049
0.0047
0.0103
***
0.1415
0.2393
-0.1278
0.3545
0.2546
0.4882
0.0714
0.0425
-0.1297
0.0483
0.0624
0.0788
0.0553
0.0447
0.1182
0.0220
0.1267
0.2573
***
***
0.1547
0.2258
-0.1098
0.2270
0.2830
0.5208
0.0414
0.0208
-0.1303
0.0485
0.0625
0.0791
0.0544
0.0450
0.1182
0.0221
0.1265
0.2595
***
***
0.1612
0.2675
-0.0891
0.3667
0.2925
0.5386
-0.0024
0.0514
-0.1784
0.0499
0.0634
0.0798
0.0564
0.0466
0.1184
0.0234
0.1277
0.2596
***
***
Import Penetration interactions
with low-skill
with semi-skill
with high-skill
Non Tradable sectors interactions
with low-skill
Electricity, gas and water
Retail & Wholesale Trade and Hotels
Transport and Communications
Financial services and other servicies to the companies
Construction
Governmental services
Personal and Household sservices
Other Services 1
Other Services 2
***
***
***
***
with semi-skill
Electricity, gas and water
Retail & Wholesale Trade and Hotels
Transport and Communications
Financial services and other servicies to the companies
Construction
Governmental services
Personal and Household sservices
Other Services 1
Other Services 2
***
***
***
*
with high-skill
Electricity, gas and water
Retail & Wholesale Trade and Hotels
Transport and Communications
Financial services and other servicies to the companies
Construction
Governmental services
Personal and Household sservices
Other Services 1
Other Services 2
***
***
***
Notes: *** if the coefficient is statistically different from zero at the one percent significance level. ** if the
coefficientis statistically different from zero at the five percent significance level. * if the coefficient is statistically
different from zero at the ten percent significance level.
31
Still, in spite of the lack of significance of the low and semi skill interactions, previous
results are maintained: import penetration increases wage inequality by increasing the
wage premium of those workers with high skills.
The results coming from the non-tradable sectors interactions are also interesting. Note that
although not significant, coefficients on low skill interactions are negative, while those
associated with high skill interactions are positive and significant in several cases, having the
semi-skilled interactions mixed results. This suggests that relative demand shifts in
nontradables have had an inequalizing impact on wage distribution.
In the case of Uruguay (see Table 11), now appears a significant negative effect from importpenetration over low-skilled workers, indicating that after controlling for shocks in nontradable output structure, import penetration have had a positive impact on wage inequality.
Still, this result is not robust to the inclusion of cluster effects in the estimation. Finally, note
that non-tradable sector interactions do not generate substantial difference across different skill
levels, being the coefficients approximately the same across skill levels for each sector. We
conclude then that for the case of Uruguay we don’t find evidence that changes in sectoral
allocation of production and employment in services has had a significant effect on relative
wages.
As indicated previously the empirical model allows the estimation of time varying
dummies corresponding to the wage premium of semi and high skilled workers relative to
the low-skilled group. Figure 7 shows the results. The reported coefficients come from
the estimation that incorporates as control variables individual characteristics as well as
the import penetration and the changes in the GDP structure. We observe that, even after
controlling by these aspects, the tertiary wage premium rise during the ’90 in both
countries, while secondary (or medium skilled) wage premium remained pretty stable.
This dynamic of high-skilled relative wages has to be associated with other factors not
directly controlled for in the regression. One key candidate is biased technological
change. If this were the case we can conclude that in Argentina this determinant accounts
32
for an increase in almost 50 percent of the wage of college graduates compared to both
low and semi skilled workers. In the case of Uruguay, during 1991-1997, technological
changed could have implied a 20% increase in the wage of incomplete and complete
college workers relative to the other two skill categories.
Figure 7: Evolution of wage premium after control the estimation by import
penetration and change in the employment structure. Argentina and Uruguay
Argentina
1.4
1.2
1
0.8
0.6
0.4
0.2
0
1992
1993
1994
1995
1996
Tertiary wage premium
1997
1998
1999
secondary school wage premium
Uruguay
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997
tertiary wage premium
33
secondary wage premiuml
6. Concluding remarks.
In this paper we combine micro-data, taking from the various household surveys
and macro data, from national accounts, to investigate the effects of different shocks in
aggregate labor demand composition on relative wages. One question that we try to
answer is: does trade liberalization play any role in shaping the Argentine and Uruguay
wage structures during the period studied? In particular, we test whether those sectors
where import penetration deepened are, ceteris paribus, the sectors where a higher
increase in wage inequality has taken place. We find evidence that supports the
hypothesis tested for Argentina but that reject it for Uruguay.
Besides trade liberalization, in these countries there has a significant expansion in the
production and employment of non tradable output. This shift in production has had significant
effects on the relative demand of workers across skill levels. Thus in the empirical analysis we
also study whether changes in labor composition within the non-tradable sector have had
significant effects on relative wages. The results for Argentina suggest that relative demand
shifts in non-tradable have had an inequalizing impact on wage distribution. In the case of
Uruguay, results are less clear cut in the sense that the identify shock in labor demand in
services do not generates substantial difference across different skill levels.
Finally a large amount of research has sought to evaluate the effect of skilledbiased technological change in wage inequality. As most of the literature (cf. e.g.
Feenstra, 1998) we indirectly identify the magnitude of this determinant through a timing
variant dummy variable for the skill variable. We obtain that this factor explains a
significant portion of the increase in wage inequality after we control for individual
characteristics, sectoral dummies, trade liberalization and shock in the composition of
labor demand across nontradables.
34
References
Acemoglu, D. (2000) “Technical Change, Inequality, and the Labor Market”, Working
paper, NBER.
Arim, R y Zoppolo, G. “Remuneraciones relativas y desigualdad en el mercado de
trabajo”. Draft.
Berman, E., Bound, J. and Machin, S. (1998): “Implications of skilled-biased
technological change: International evidence”, Quarterly Journal of Economics, vol. 113,
pp. 1245-80.
Bernard, A., Bradford,J. (1998). “Understanding Increasing and Decreasing Wage
Inequality”, Working paper, NBER.
Bound, J., and Johnson, G. (1992): “Changes in the structure of wages in the 1980s: An
evaluation of alternative explanations”, American Economic Review, vol. 82, pp. 371-92.
Casacuberta, C y Vaillant, M. (2001) “Trade and jobs in Uruguay ‘90”. Draft.
Feenstra, R. (1998): “Integration of trade and disintegration of production in the global
economy”, Journal of Economic Perspectives, vol. 12, pp. 31-50.
Feenstra, R. (2000): The impact of international trade on wages, NBER, University of
Chicago Press.
Galiani, S. (1999): “The differential evolution of wages, job stability and
unemployment”, mimeo.
Galiani, S and Sanguinetti. (2000): “Wage Inequality and Trade Liberalization: Evidence
from Argentina”
Katz, L. and Murphy, M. (1992): “Changes in relative wages, 1963-1987: Supply and
demand factors”, Quarterly Journal of Economics, vol. 107, pp. 35-78.
Krugman, P. (2000): “An alternative model of trade, education and inequality”, in The
impact of international trade on wages, Feenstra, R. (ed.), NBER, University of Chicago
Press.
Kydland, F and Zarazaga, C.(1997: “Is the business cycle of the Argentina different?.
Economic Review, Federal Reserve Bank of Dallas.
Lawrence, R. and Slaughter, M. (1993): “Trade and US wages: Giant sucking sound or
small hiccup? Brookings Papers on Economic Activity, pp. 161-226.
35
Leamer, E. (1998): “In search of Stolper-Samuelson linkages between international trade
and lower wages”, in Imports, exports, and the American worker, Collins, S. (ed.),
Brookings Institution.
Richardson, J. (1995): “Income inequality and trade: How to think, what to conclude”,
Journal of Economic Perspectives, vol. 9, pp. 33-55.
Wood, A. (1995): “How trade hurt unskilled workers”, Journal of Economic
Perspectives, vol. 9, pp. 57-80.
36