NTC Project: S03-PH02
1
Project No.: S03-PH02
Principal Investigator: Anusua Datta, Ph.D., Philadelphia Univ., Industry Analysis/Econometrics
Email: dattaa@philau.edu
; Phone: (215)-951-2916
Members: Susan Christoffersen, Ph.D., Philadelphia Univ., Economics/Industrial Organization.
D.K. Malhotra, Ph.D., Philadelphia Univ., Financial Analysis
Goals:
Philip Russel, Ph.D., Philadelphia Univ., Quantitative Analysis
Mikhail Kouliavtsev, Ph.D., Philadelphia Univ., International Trade
Helmut H.A. Hergeth, Ph.D., N.C. State Univ., Cost Management in Textiles
The goal of this research is to identify ways in which to improve the competitiveness of U.S. textiles and apparel industries. To do so: 1. We examine the cost structure of the two sectors to determine the least cost methods of production; 2. Consider financing decisions of firms as a means to maximize shareholder wealth; 3. We examine the impact of labor wages, tariffs, exchange rates and NAFTA on the pattern of
U.S. textile trade with the rest of the world.
Abstract
At the industry level, we use a translog cost function model to assess the underlying technological factors that affect efficiency and cost in the U.S. textile and apparel industries. We estimate substitution elasticities between inputs (labor, capital, energy and non-energy materials), study the nature of technical change and determine the economies of scale in both industries. This analysis is then extended to study
Non-wovens, a sector within textiles that has experienced fast growth in recent years largely due to product and process innovation.. Global competition from cheap imports has been a major factor the U.S. textiles industry has had to contend with. Following recent literature in trade theory, we utilize a partialequilibrium gravity model to analyze the pattern of U.S. textile and apparel imports as a function of country wages, tariffs rates and fluctuations in exchange rates at the industry and 4-digit sector level, using bilateral trade data for 65 countries from 1989 to 2001. Further we consider the trade-creation and trade-diversion effects of NAFTA. Finally, we use firm level data to look at the financing decisions (debt and equity mix) of textile and apparel firms, which have a direct bearing on firms’ economic value and profitability.
Overview
The U.S. textile and apparel industries have faced difficult times over the past three decades. This sector has seen large-scale downsizing, with the share of manufacturing employment declining from 12.1% the
1970s to 8.1% in the 1990s (Francisco, 2000). Intense import competition from low wage developing countries in apparel and both developed and developing countries in textiles has been a major factor to contend with. Competition has intensified with the gradual phasing out of textile and apparel quotas.
The crisis facing the domestic textile and apparel industries requires a broad based analysis of the factors that impact the competitiveness of the sector. This research is dedicated to accomplish this mission. At the
industry level a cost (profit) function model is used to quantify substitution possibilities among inputs in order to find the optimal input mix, study the nature of technological change and its role in improving competitiveness, and determine optimal “firm size.” Sector-wise data on bilateral trade in textiles is used to assess the impact of tariffs, exchange rates, wages, and regional trade agreements on U.S. textile imports.
National Textile Center Annual Report: November 2006

NTC Project: S03-PH02
2
In the face of global competition, textile firms have lost economic value in the market with a decline in the prices of textile stocks. We use firm-level data for the U.S. textile industry for the period 1987 to 2000 to consider the determinants of firm capital structure, which have a significant bearing on shareholder wealth.
Summary of Results:
Cost statistics, spanning the period 1949-1999, indicate that labor and material costs account for the highest share in textile and apparel production costs. The scope for factor substitution in textiles remains limited, with all substitution elasticities less than 0.5. Labor and materials are complements in apparel production, but there are substitution possibilities between capital and labor. In both industries technical progress is labor saving. The rate of technological progress is higher in textiles than in apparel. Overall economies of scale are larger in apparel, suggesting larger firms enjoy lower costs. Scale economies show an upward trend in textiles, indicating a move towards larger firms.
We examined the impact of wages, tariffs, exchange rates and NAFTA – on the pattern of U.S. textile and apparel trade at the industry and 4-digit sector level. The results indicate that bilateral tariffs and exchange rates largely determine a country’s share in U.S. textile imports, while wages are a significant factor in apparel trade. NAFTA had a significant trade creating effect with Mexico and Canada, whose share of U.S. textile imports increased significantly after 1994. Evidence of trade diversion from other trading partners to the NAFTA members is not very strong, with only imports from Latin America showing a significant decline since the passage of NAFTA. Sector analysis shows a lot of variation.
Our firm-level analysis indicates that with respect to financing decisions, textile and apparel firms seem to finance capital expenditure more through equity offerings rather than through borrowings. Higher tax rates and tax-shields encourage more borrowing. Finally, larger firms have a higher debt-equity ratio than smaller ones.
Part I: Industry Level Analysis: Cost Structure and Factor Use in Textiles and Apparel
A.
Input and Output Prices and Factor Cost Shares
Over the past 50 years prices of textile products have increased very modestly. As Figure I.1 indicates, output prices rose by about 54%. In comparison input prices rose at a much faster pace. The fastest growing component of cost is labor wages. Real wage rates between 1949 and 1999 increased by as much as ten folds. The price of energy, raw material, and capital rental prices increased between three to five folds. Output prices have remained low, due largely to competition from low cost imports.
Figure I.1: Trends in Output and Input Prices in Textiles and Apparel Industries
Apparel
Textile
12.00
10.00
8.00
6.00
4.00
2.00
0.00
1953 1961 1969 1977 1985 1993 2001
Year
10
8
6
4
2
0
1953 1960 1967 1974 1981 1988 1995
Year
Output Price
Capital
Service
Price
Real Wages
Energy Price
Material
Price
National Textile Center Annual Report: November 2006

NTC Project: S03-PH02
3
The apparel industry shows similar trends. In apparel labor wages increased by a factor of eight, while output prices merely doubled between 1953 and 2001. The growth in the price of capital, energy and materials in apparel ranged between 1.7 to 3 folds.
Cost Shares: Labor accounted for the biggest share of textile production costs before 1975. The share of labor, which was nearly one-half of total costs in the early 1950s, gradually declined to about one-third by the end of the 1990s. Share of materials increased from about 30% to a little more than 40%. The share of capital in total costs fluctuated between 10-15%, while energy accounted for less than 7% of total cost. The production structure of the U.S. textile industry indicates that it is still fairly labor intensive.
As expected, apparel production is highly labor and material intensive. Raw materials account for 50% -
60% of total cost, while the share of labor is between 30% and 43%. The share of labor declined from a high of 44% in 1970 to below 30% in 2001. The share of capital on the other hand, which was below 5% at the beginning of the sample period, increased steadily to around 10% by the end of 2001.
Energy costs have remained roughly around 1% of costs in apparel.
B. Translog Cost Function Estimation
Assuming firms minimize total costs of production, the general form of the aggregate cost function is expressed in its translog form, a second-order approximation to an arbitrary twice-differentiable surface
[Christensen et al. (1973)], as l n C
 
0



  q i ln Q
  
P i
T
 i

 i ln P i
 1
2
 
 t qq

(ln
1
2
Q

)
2  tt
T
1
2
 i
 j
 ij ln t ln qt ln QT T
2 where i,j = K,L,E,M, and

are the parameters to be estimated.
P i ln P j
  i
 iq ln P i ln Q
(1)
An important objective of this study is to identify the sources of cost savings that can allow the U.S. textile and apparel industries to remain competitive. Cost savings can occur through scale economies, represented by movements along the average cost curve, technical change, associated with the downward displacement of the unit cost curve, and the factor bias of technical change, which alters the optimal level and mix of inputs.
Scale economies (

Q
) are measured directly as the reciprocal of elasticity of cost (

CQ
) with respect to output 1 ,

Q
=1 characterizes constant returns to scale.

Q
>1 or

Q
<1 imply increasing or decreasing returns to scale. The rate of technical change (

T
) equals the negative of the rate of growth of total cost with respect to time, holding output and prices of all inputs constant 2 . In equation (2),

' s measure the biases in technical progress. Technical change is ith factor saving if
 it

0 and factor using if
 it

0 . For example,

LT

0 and

KT

0 would indicate labor-saving and capital using technical progress. The parameters
 and

, measure neutral technical change, characterized by pure shifts in the cost t tt function. The Allen-Uzawa partial elasticities of substitution between two factors i and j,

ij, are computed
1

Q



 ln
 ln
C
Q



1


 q

  iq ln P i
  qq ln Q
  i
T


1
; 2

T  
 ln

T
C
 

 t
   it ln P i
  qt ln Q
  tt
T
 ; 3  ij

 ij
 S i
Sj
S i
S j
National Textile Center Annual Report: November 2006

NTC Project: S03-PH02
4 directly from the translog cost function 3 . A positive value of

ij indicates inputs i and j are substitutes while a negative value indicates they are complements. Moreover

ij closer to 1 indicates inputs i and j are highly substitutable, while a value close to 0 indicates low substitutability.
This study is based on data for the period 1953-2001. Data on cost and prices of labor, capital service, energy, non-energy materials and real output for textiles (SIC 22) and apparel (SIC 23) are taken from the
Bureau of Labor Statistics’ Multifactor Productivity database.
The production-cost structure of the U.S. textile and apparel industries was examined using a translog cost framework. In Textiles, the average elasticity of substitution between capital and labor (
 lk =0.34) is significantly below one. This suggests that the scope for substitution between capital and labor remains limited in textiles. Similarly, the magnitude of substitution elasticities between labor-materials, capitalmaterials and capital-energy are all less than 0.5, indicating low overall substitutability among all inputs
Table I.1: Allen-Uzawa Elasticities of Substitution and Technical Change (Average: 1953-2001)
Textiles
Apparel

lk
0.338
1.182

lm
0.276
-0.395

le
0.244
0.776

ke
0.246
1.768

em
-0.501
0.027

mk
0.115
-0.359

T
2.4%
0.7%

Q
1.00
1.16 in textile production. In Apparel, labor and materials are complements (
 lm = - 0.39), as indicated by a negative coefficient, but there are greater substitution possibilities between labor-capital
 lk =1.18) and capital-energy and (
 ke = 1.768) suggesting more capital intensive production and greater fuel efficiency.
Annual cost reductions from technical change (

T ) have averaged 2.4% per year in textiles and about 0.7% per year in apparel. In both industries technical change is labor-saving and capital-using. This suggests that the effect of increased competition from low-wage countries has forced rational producers to shift away from the use of the relatively more expensive labor input. Scale economies (

Q
) in textiles rose steadily from 0.82 (

Q
<1,indicating decreasing returns), to 1.21 between 1998-2001, indicating a 0.21% decrease in total cost for every 1% increase in output. This suggests a larger firm size is more conducive for reducing costs. The average scale economies in apparel were higher, at about 1.16 for the period 1953-
2001 .
Non-Wovens: A New Paradigm for Success?
A translog cost function can also be constructed for a subset of the industry. Nonwoven fabric mills are a growing sector of the textile industry; what was once a niche market may now exemplify the future of textiles. The nonwoven sector shipped a total of $5.8 billion dollars in 2002 (Census Bureau, 2004). In 2002 the sector employed more than 21,000 employees (Census, 2004) at average wage rates above the typical textile worker. The growth trends begin in the early 1970’s, with the introduction of product and process innovations. The challenge that faces all textile manufacturers is the prices of inputs have been steadily increasing, outpacing increases in the final prices for output. Fig. I.1 above highlights these price pressures.
While these profiles are common within textile manufacturing, there are two substantial differences in the nonwovens sector. First, the relative capital intensity of the nonwovens sector protects it from
National Textile Center Annual Report: November 2006

NTC Project: S03-PH02
5 much of the competition from developing countries: A second important difference is that the cost of material closely tracks the price index for shipments (the final price of the product) which diverges from the general pattern for textile sectors where we find material costs increasing by much more than output prices. The ability to match output price to material input price increases, as well as immunity to foreign pressure, both contribute to the continued viability in this sector.
We apply econometric techniques to analyze this sector’s success, particularly with regard to changing patterns of input use, that is, the degree of substitutability between inputs as costs change, as well determine the economies of scale. The estimates are based on the model outlined in Part I.
Figure II.1: Economies of Scale Figure II.2: Technical Change
2.5
2
1.5
1
0.5
0
1958 1963 1968 1973 1978 1983 1988 1993
Year
0.04
0.03
0.02
0.01
0.00
-0.01
1958 1963 1968 1973 1978 1983 1988 1993
-0.02
-0.03
Year
Decreasing returns to scale are indicated by

Q
< 1, that is, costs increase more than proportionate to the increase in output. This is what we observe in the nonwoven sector up until the 1970’s at which time that sector began a period of notable growth. After 1975, we see a transformation;

Q
>1, implying increasing returns to scale. The industry became more efficient between 1978 and 1996, as output grew. This is consistent with a capital intensive industry, requiring large fixed costs. As demand for the products grows, labor is added to take advantage of the scale economies. For this sector we find

KT

0 and

ET

0 which indicate capital-saving and energy using technical progress. Again we see a shift in technology taking place around 1970; the energy using component may be attributed to the increased fuel prices at the time. The parameters
 and t
 tt
measure neutral shifts in the cost function. As both are negative and significant, we find costs are going down over time, at a decreasing rate.
Elasticities of Substitution
Substitution elasticities between energy and non-energy inputs are given in Fig. II. 1 and 2. A positive elasticity indicates that the inputs are substitutes, a negative estimate indicates complements. Previous to
1974, if one were to increase production, using more labor or material, one would necessarily use more energy. After 1974, one could substitute more capital or material for energy, perhaps indicating
Fig. II.1 Substitution Elasticities: Energy Fig II.2 Substitution Elasticities: Non-Energy
1.00
0.50
0.00
-0.50
1958 1963 1968 1973 1978 1983 1988 1993
-1.00
-1.50
Energy-Material
Year
Energy-Labor Energy-Capital
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
1958 1964 1970 1976 1982 1988 1994
Year
Material-Labor Material-Capital Labor-Capital
National Textile Center Annual Report: November 2006

NTC Project: S03-PH02
6 new energy saving capital investment. While material and capital remain highly substitutable, material and labor decline in the elasticity of substitution in production. This period covers a technological transition to greater degrees of capital intensity in production. As fewer workers are involved with the production process, they are more critical and less substitutable for other inputs.
The high degree of substitutability between materials and output might at first blush seem counterintuitive; how can one use more machines and less fiber to achieve a certain level of production?
Could one produce the same amount of sweaters, for example, with more looms and less wool? The degree to which capital and material are not complements perhaps reflects the ability of the industry to outsource; that is “to make or buy”. If they are making the intermediate inputs, they invest in capital; if the prices shift, an agile manufacturer buys the intermediate inputs, thereby substituting material for capital.
The purpose of this study is to investigate the patterns of the United States textile trade with the rest of the world. Specifically, we are interested in the composition of the imports of textile goods into the U.S. and how it changes over time in response to various shocks.
Figures III.1 and III.2 give the textile import shares and tariff rates by region. Asia accounts for the highest share of U.S. textile imports at around 58%. Between 1989 and 2000 the imports from Western
Europe declined from 30% to 15%, while that of NAFTA and Caribbean countries increased from 3%-14% and 1%-7% respectively. The figure on tariffs indicates that lowering of tariff rates between the U.S. and
Nafta and Caribbean countries has played a significant role in increasing U.S. trade with the two regions.
Figure III.1: Textiles: Import Share by Region Figure III.2: Textiles- Tariff Rates by Region
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1989 1991 1993 1995 1997 1999
Year
0.2
0.15
0.1
0.05
0
1989 1992 1995
Year
1998
Asia
Caribbean
W. Europe
Latin
NAFTA
Following recent literature in trade theory, we modify the a partial-equilibrium gravity model proposed by Fukao, Okubo and Stern (2003) to analyze the impact of wages, tariffs rates, fluctuations in exchange rates and NAFTA on textile and apparel imports at the sector level. Further we consider the trade-creation and trade-diversion effects of NAFTA. Additional regional dummy variables for exchange rate and per capita GDP are included to study their differential impact on U.S. imports by region. ln s it
=

i
+
 t
+

1 ln wage it
+

2 ln Tariff it
+

3 ln Exchange
it
+

1
NAFTA it
+

2
AFRICA it
+

3
ASIA it
+

 WEUR it
+

 CARIB +

6
LATIN +

7
MENA it
+

8
OCEAN it
+
 it
(2)
where i indexes countries; the dependent variable is the natural logarithm of each country’s import share
(s it
) in the U.S. market. wage it
is the labor wage in country i in year t. Tariff it is one plus the effective tariff imposed by the U.S. on the applicable imported goods from country i, and Exchange
it
is the real exchange
National Textile Center Annual Report: November 2006

NTC Project: S03-PH02
7 rate.
 

 are the region dummies that capture the effect of NAFTA on countries in the NAFTA (U.S.,
Mexico, Canada), Africa, Asia, Western Europe, Caribbean, Latin America, Middle East and North Africa and Oceania. Bilateral trade data for 65 countries over the period 1989-2001 and at the 4-digit SIC sector level has been collected for the purpose for both textile and apparel industries. The data on bilateral U.S. trade in textiles are taken from the US Bilateral Manufacturing Imports and Exports by SIC4 (1987 revision),
1989 to 2001 database (Feenstra, Romalis, and Schott 2002).
Table III.1 Dependent Variable: Import Share
Constant
RGDPPC
RXRATE
TARIFF
NAFTA
NAF_AFRICA
NAF_ASIA
NAF_CARIBB
NAF_LATIN
NAF_WEUR
NAF_MENA
NAF_OCEANIA
Model 1
2-way FEM
-8.0379***
0.2277**
0.0907***
-2.5867**
0.9958***
Adj R 2 = 0.8867
N= 735
TEXTILES
Model 3
2-way FEM
-8.4762***
0.2667**
0.0955***
-2.6432**
0.9033**
-0.3354
Model 1
2-way FEM
7.3114**
-1.5828***
-0.0372
-0.0579
0.3642
-0.0145
0.0654
-0.5203**
-0.1696
0.1429
-0.8971*
Adj R 2 = 0.8871 Adj R 2 = 0.9464
N = 771
APPAREL
Model 3
2-way FEM
-8.5503***
0.2329
-0.0399
-0.031
0.5931*
0.4133**
-0.0802
0.5525**
-0.1903**
-0.1771
0.4842*
0.5997
Adj R 2 = 0.9463
* Significant at the .10 level, ** significant at the .05 level, ***significant at the .01 level.
The results for the textile industry are presented in columns 1 and 2 of Table III.1. In Model 1, tariff and exchange rate measures are significant and have the expected signs: a depreciation of country i’s currency against the U.S. dollar is associated with an increase in i’s share of U.S. textile imports, 4 while an increase in the effective tariff rate against country i’s textile output lowers that country’s share. The coefficient on
NAFTA is positive and significant, indicating that the shares of Canada and Mexico increased after 1994.
The somewhat surprising result is the positive coefficient of real GDP per capita: one would expect a negative relationship between the country’s prevailing wage rate and its importance in exporting textiles
(a relatively labor-intensive good) to the U.S. It appears that low wages are not a significant factor in textile trade.
Trade Creation and Diversion: To investigate the impact of NAFTA further, we introduce regional dummy variables, as described in equation (2). This allows us to determine whether the shares of U.S.’s NAFTA trading partners (Mexico and Canada) increased at the expense of countries in other regions. The results are presented under Model 2. The explanatory variables have the same signs and similar magnitudes, relative to the simpler model discussed above, which provides some reassurance about the robustness of our approach. The addition of regional NAFTA measures, indicates that the NAFTA variable is still significantly positive confirming the boost in Canadian and Mexican import shares, while NAF-LATIN has a significant negative coefficient indicating that countries of Latin America saw a decrease in their
4 Exchange rates are in terms of units of country i’s currency for one U.S. dollar.
National Textile Center Annual Report: November 2006

NTC Project: S03-PH02
8 shares of U.S. textile imports after 1994. The main result of interest in this model is the insignificance of the NAF-ASIA’s coefficient. It appears that NAFTA did not increase the shares of Mexico and Canada at the expense of Asian countries’ textiles. In other words, we find no significant evidence of trade diversion from Asia attributable to the passage of NAFTA.
The results for the apparel industry are presented in columns 3 and 4 of Table III.1. In contrast with the textile industry, we find that our wage measure has the most explanatory power of all of the right-handside variables in the simpler models. It appears that differences (and changes) in wages explain all of the variation in the shares of apparel imports. Adding the regional NAFTA indicators leaves all other variables, including RGDPC insignificant. The positive coefficients on NAF-AFRICA, NAF-ASIA, NAF-
MENA, and NAFTA suggest that countries in these regions saw their shares of U.S. apparel imports increase after 1994. However, the insignificance of our regressors indicates that sector analysis may reveal some within-industry differences at the more disaggregated level.
We estimate our simplest model for each of the 15 four-digit SIC textile sectors and 27 apparel sectors in our sample. The results are rather encouraging in that the signs and magnitudes of our estimates mirror those we obtain for the textile industry as a whole.
Textile: The coefficient on our measure of tariffs is negative for all sectors and statistically significant for all but three (Broadwovens – manmade fiber and silk, Thread Mills, and Textile Goods not elsewhere classified). The real exchange rate has a significantly positive effect on country i’s share of imports into the U.S. for all but four of the sectors (Carpets and Rugs, Thread Mills, Coated Fabrics, not rubberized, and Tire and Cord fabrics). This result is consistent with the relatively low volume of trade in these sectors (relative to other sectors); Only three textile sectors saw any significant gain in the shares of Mexico and Canada after 1994: Hosiery, not elsewhere classified, Thread Mills, and Tire Cord and fabrics.
Interestingly, Thread Mills and Tire Cord sectors are both characterized as relatively capital intensive and experiencing fast growth in the 1990s. Addition of the other NAFTA regional dummies (Model 2) reveals that trade diversion (i.e. diversion of trade from the rest-of-the-world to North America) especially from
Asia is negligible.
Apparel: For apparel sectors, the effects of tariffs and exchange rates are consistent with our theoretical predictions: higher values of real exchange rates (i.e., stronger U.S. dollar) are generally associated with higher shares of imports, while higher effective tariffs correlate with lower import shares. Wages have a significantly negative effect for eight sectors in our parsimonious specification. We conclude that the effect of wages in our models is stronger than in our similar estimates for textiles; this result appears in line with the stylized fact that apparel production is more labor intensive. Our primary variable of interest,
NAFTA, behaves as expected: it is significant for 16 sectors and is positive for all of them, suggesting that
Mexico and Canada saw their shares of U.S. imports of these goods increase after 1993. The sectors for which NAFTA appears to have no effect are generally the labor-intensive sectors exhibiting slow growth and low levels of R&D spending: Based on our results, we conclude that in general there is no evidence of
trade diversion from Asia to NAFTA countries. In fact, in some sectors of the apparel industry, Asian countries as a group saw their shares of U.S. imports increase in 1994 and thereafter. On the other hand, some trade it appears was diverted from Western Europe as we found these countries’ shares declined for many sectors.
National Textile Center Annual Report: November 2006

NTC Project: S03-PH02
9
In this study, we analyze the capital structure decisions of textile and apparel firms. Capital structure decisions are critical and reflect a firm’s strategy with regard to financing real investments. A survey of modern corporate finance literature shows that capital structure decision is an important part of shareholder wealth maximization principle. We model the capital structure as a function of capital expenditure, size of the firm, profitability, corporate tax rate, growth in sales, non-depreciation tax shield and import penetration. Using the data for 22 textiles and 22 apparel firms for the period 1987 to 2000, we use panel data approach to analyze the factors that determine the debt-equity ratio in this sector.
Table IV.1 and IV.2, report the results from the estimation of the debt-equity ratio for U.S. textile and apparel firms respectively. For textile firms (Table IV.1) coefficient for profitability (PROF) is negative, which indicates that a profitable firm favors internal sources of financing consistent with pecking order theory. Firm size (ASSET) and growth in sales (GROWTH) have the expected positive and negative signs respectively. Larger textile firms, tend to carry a higher debt relative to smaller firms. Also, firms exhibiting growth in sales revenue, tend to rely less on borrowing. The signs on the tax variables (tax rate and non-debt tax shield) are as expected although the coefficients are fairly small suggesting that these variables are not economically significant.
In Model 2 we study the impact of import penetration. With the addition of the penetration variable, the adjusted-R square increases. Capital expenditure, size, profit and growth remain significant while Tax rate and NDTS are now insignificant. Note that both tax rate and NDTS has very small coefficients in model 1. The coefficient for import penetration, the variable of interest, is negative. This suggests that increased import competition causes textile firms to take on lower amounts of debt, perhaps to lower business risk.
Table IV. 1: Textile Industry Table IV.2: Apparel Industry
Dependent Variable: Debt-Equity Ratio
Variable Model 1
CAPEXP
SIZE
-0.883
0.180
**
***
Model 2
-1.03
***
0.145
***
Dependent Variable: Debt-Equity Ratio
Variable Model 1
CAPEXP
SIZE
-0.368
0.008
PROF
GRTH
-1.244
***
-0.165
***
-1.089
**
-0.137
*
PROF
GRTH
NDTS
TAX
-0.003
***
0.001
**
-0.0017
-0.0107
NDTS
TAX
PEN -1.661
* PEN
R 2 (adjusted) 0.538 0.593 R 2 (adjusted)
N 209 136 N
*significant at .10 level; ** significant at .05 level; *** significant at 0.01 level
-0.543
**
-0.022
0.000
-0.145
**
0.569
281
Model 2
-0.0002
-0.074
***
-0.517
*
-0.135
***
-0.000
0.0237
0.0002
*
0.730
182
For the apparel industry, only two variables, profitability (PROF) and tax rate (TAX), are statistically significant in model 1. Profitability is negatively related to the debt-equity ratio indicating that more profitable firms use less debt in their capital structure. Surprisingly, the corporate tax rate has a negative coefficient implying that if the corporate tax rate is high, apparel firms borrow less. This may be due to the fact that apparel firms do not have enough operating profits to benefit from tax write-offs for interest payment on corporate debt. In model 2, we add import penetration (PEN) to study the effects of cheap imports on financial leverage. The parameter estimates for firm size (Size) and growth in sales (Growth) and import penetration (PEN) are significant. Surprisingly, firm size has a negative sign suggesting that larger apparel firms tend to borrow less. The negative sign for growth confirms firm’s preference for internal sources of financing. The coefficient of PEN is positive suggesting that when import competition
National Textile Center Annual Report: November 2006

NTC Project: S03-PH02
10 increases apparel firms tend to borrow more. This result is the opposite of what we find for the textile industry. Thus apparel firms seem to respond to the challenge of cheaper imports by borrowing heavily to finance investment in technology to improve productivity.
Future Direction:
We have produced 5 papers from this project of which 2 have been published. The other 3 papers are under review at journals like the Review of International Economics, Journal of Economics & Business and The
Oxford Journal. The results of our research have been disseminated at various national and international conferences. Many researchers from U.S. and foreign universities have shown a significant interest in this work, as evidenced from the number of email queries I have received. We hope to publish a book with all the findings from our research.
Contributing Graduate Students:
Hitesh Khurana, MS Textile Engg, Peeyal Banerjee, MBA (Fin) Philadelphia University, Saqib Sohail, NCSU
Interactions: Academic 1[UC-Davis]; Industry 8 [ATMI, OTEXA, Burlington, Liz Claiborne, P/Kaufman
UNIFI, Pacific Century Textiles (HK) Ltd.
]
Scholarly Publications:
1. Datta, A. and S. Christoffersen. 2005. "Factor Substitution, Returns to Scale and Technical Change in the U.S.
Textile and Apparel Industries." Atlantic Economic Journal, 32, pp. 1-13.
2. Datta, A, D.K. Malhotra and P. Russel. 2006. "Regional Trade Pacts and the Competitiveness of the U.S. Textile
Industry." Forthcoming in Competitiveness Review, Vol. 16.
3. Datta, A., D.K. Malhotra and P. Russel. 2006. "Determinants of Capital Structure: An Empirical Analysis of the
U.S. Textile and Apparel Industries.” (under review)
4. Datta A. and M. Kouliavtsev. 2006. "NAFTA and the Realignment of U.S. Textile and Apparel Trade: Trade
Creation or Trade Diversion?” (under review)
5. Datta, A. and S. Christoffersen. 2005.
“The Nonwoven Textile Sector: Elasticities of Scale and Substitution.”
(under review)
Conference Papers and Proceedings:
1. Proceedings of the Global Conference on Business and Economics, July 2004. "Factor Substitution, Scale Economies and
Technical Change in the U.S. Textile and Apparel Industries."
2. Midwest Economics Association Meeting, Milwaukee, March, 2005: "The cost structure of U.S. Textile and Apparel
Industries: A Sector Analysis." (with S. Christoffersen)
3. Eastern Economic Association Conference, New York, March, 2005: "NAFTA and the Realignment of U.S. Textile
Trade: Trade Creation or Trade Diversion?" (with M. Kouliavtsev).
4. Proceedings of the Global Conference on Business and Economics, July 2006. "Determinants of Capital Structure: An
Empirical Analysis of the U.S. Textile and Apparel Industries.” (with D.K. Malhotra and P. Russel)
5. Centre for Economic and Social Studies, Hyderabad, India, Aug 2006,"NAFTA and the Realignment of U.S. Textile and
Apparel Trade: Trade Creation or Trade Diversion?” (with Mikhail Kouliavtsev)
URL:
References:
Christensen L. R., D. W. Jorgensen and L. J. Lau. 1973. Transcendental Logarithmic Production Frontiers. Review of Economics and Statistics ¸ vol 55 (1), pp 28-45.
Fukao K., T. Okubo, and R.M. Stern (2003). An Econometric Analysis of Trade Diversion under NAFTA.
Discussion Paper Series A, No. 431, Institute of Economic Research, Hitotsubashi University, Japan.
Harris, M and A. Raviv (1991) “The Theory of Capital Structure,” Journal of Finance, 46: 297-355 .
National Textile Center Annual Report: November 2006

NTC Project: S03-PH02
11
Hufbauer, G. and J. Schott. 1993. NAFTA: An Assessment. Washington D.C.: Institute of International Economics.
Ramolis, John. 2002. “NAFTA’s and CUSTFA’s Impact on North American Trade.” Univ.of Chicago Working
Paper.
Myers, S. C. (2001) “Capital Structure,” Journal of Economic Perspectives , 15: 81-103.
ATMI, 2001. Crisis in U.S. Textiles. Office of the Chief Economist and the International Trade Division of the
American Textile Manufacturers Institute, August, Report.
National Textile Center Annual Report: November 2006