MULTINATIONAL CORPORATIONS IN LDCs: A COMMENT
| DOI | http://doi.org/10.1111/j.1468-0084.1977.mp39003007.x |
| Date | 01 August 1977 |
| Author | DAVID E. DE MEZA |
| Published date | 01 August 1977 |
MULTINATIONAL CORPORATIONS IN LDCs:
A COMMENT
By DAVID E. DE MEZA*
I
In a recent paper in the BULLETIN Courtney and Leipziger1 (henceforth CL)
undertake an empirical investigation of 'the extent to which MNCs transfer a
standardized technology to their affiliates in DCs and LDCs and the extent to
which these technologies are modified by local wage-interest constraints, ex ante or
ex post' (pp. 298). They conclude 'we have presented the case that MNCs do
respond to factor prices in their choice of technology' (pp. 303) and that 'LDC host
governments would do well to cease subsidization of capital costs relative to wage
costs if they wish to stimulate employment' (pp. 302). However, the methodology
applied is based on the estimation of aggregate production functions. The con-
ditions which must be met for this to represent a valid procedure are stringent and
both the internal evidence of the paper and the results of other studies suggest they
have not been fulfilled.
Using cross-section data for US majority-owned manufacturing affiliates, C-L
estimate for each of II two and three digit industries a separate Cobb-Douglas
production function for LDCs and DCs. Value added is regressed on total capital
assets and labour employed, and it is found that:
42 out of 44 input-output elasticities are statistically significant;
for all equations R2 exceeds 0.78;
a constant returns hypothesis cannot be rejected in 17 out of 22 cases;
in 5 out of 11 industries the production function does not differ significantly
between LDCs and DCs.
In Part II of this comment some theoretical difficulties involved in interpreting
these results will be outlined, and in Part III an attempt will be made to discover
the practical importance of the points raised.
II
It will initially be assumed that each industry contains a single MNC producing
just one good by means of homogeneous labour and capital. A conventional pro-
duction function which relates the physical flows of industry output to the real
flows of capital and labour inputs can then be defined. Whilst it will subsequently
be argued that the data are not at all consistent with these assumptions, this is
nevertheless the case in which the conclusions of C-L appear most reasonable.
Even so, in the cross-section regressions performed by C-L, the value of output is
regressed on the value of capital and the number of workers employed. As long as
* I am grateful to J. Creedy, J. J. Thomas, and the Editors for helpful comments. Errors
are of course my own responsibility.
Courtney, W. H. and Leipziger D. M. 'Multinational Corporations in LDCs: The Choice of
Technology', BULLETIN, Nov. 1975, pp. 297-303.
To continue reading
Request your trial