The value of international outsourcing: a study on Veneto clothing industry Carlo Gianelle, University of Siena, Doctoral School Giuseppe Tattara, University of Venice, Dept. of Economics October 2007
In the 80s many production phases have been delegated to Italian
outworkers (cutting, dyeing, sewing, stitching and pressing – apparel,
stitching – footwear)
Phases at the beginning or end of the production chain require human
capital (creation, modelling etc.) and sophisticated machinery (cutting,
washing, dyeing and printing -> apparel, moulded soles -> footwear)
The number of employees working in SMEs textile-clothing-footwear
in Veneto sharply increased in the 70s and declined subsequently in
the 90s because a vast majority of subcontractors are foreign
(foreign outsourcing -> Benetton is an earlier example: from 80%
domestic to 20% domestic in 5 years)
Veneto: apparel and footwear 0 10000 20000 30000 40000 50000 60000 70000 80000 90000 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Number of employees Artisan firms Big firms not-artisan TOT
We consider outsourcing abroad as a new policy adopted by the final
firms and try to evaluate its effect at the firm level
Does the decision of outsourcing abroad lead to:
Expansion of the activity in terms of Turnover?
N ecessary to defend or increase the market share
Realization of more profits in terms of Gross earnings (Ebitda)?
‘ Lean and mean’ : anecdotal evidence -> final firms (brands) slice the production process iff they are able to embed
a larger share of profits
The effects of outsourcing
Did delocalization increase Turnover and Gross profits?
Is the effect higher, the higher the quota of products realized abroad?
Was this a lasting improvement or a once for all bonus?
Fixed-effect impact equation using panel data referring to final producers
logY is alternately the Turnover and the Gross earning (EBITDA)
both expressed in logarithms
Impact of outsourcing is estimated by means of the dummy Dc that splits the time series in 2 periods: before and after the relocation event
Linear trend T also included. Dummy Dc interacts with T , resulting in the variable TDc that captures the growth effect of relocation
Controls: sector orders at the international level Ord ; year dummies γ t
Self-selected group of 70 limited and joint stock companies based in Veneto (1/4 of the total employment of companies with more than 50 employees ), final producers in the clothing and footwear sector, which have delocalized some important production phases
Mostly medium-size firms , with average workforce of 110, average turnover of 36 millions of Euros
Budget data from the Veneto Provincial Chamber of Commerce collection; employment data from the VWH database; data on outsourcing from a questionnaire delivered to each firm and supplemented by several telephone interviews
Time: 1982 – 2003, unbalanced panel; 16 obs. per firm on average
Countries of delocalization: Romania , China , Tunisia , Hungary, Bulgaria, Croatia, Turkey, Slovakia, India, Indonesia, etc.
Average and growth effect of relocation
Average effect : drift of the continuous line
Growth effect : difference in the dotted line slope
t d =0 time EBITDA, Turnover
Effects of relocation: visual representation
Resid. of regress. of logTurn (left) and logEBITDA (right) over controls
In the case of EBITDA there’s a DIP : some firms delocalize after a
drop in gross (net) profits -> reasonable result but caution is required
in interpreting the coefficient estimates: firms self-select into treatment
Problem: the average effect of relocation tend to be overestimated if