A comment on the following dissertation suggestion by Duncan:
>2. The publication of the Penn World Tables data on comparative growth
>patterns since 1960 for a large sample of countries opens up the
>possibility of some systematic examination of issues related to the falling
>rate of profit (such as changes in the rate of exploitation and the
>composition of costs and of capital) empirically.
>
Based on my experience with estimates of the rate of profit in Mexico,
I would be very careful using sources such as this to estimate the rate of
profit and the rate of exploitation, especially in developing countries.
Actually, it would be very interesting to study carefully their methods of
estimation. But in general the data are most less reliable and have to be
used with great caution. For example, the capital stock data, which are
based on an annual survey rather than the perpetual inventory method.
Another big problem - at least in Mexico - is that there are no separate,
independent estimates of profit (hence no "income" or "value added" approach
to the estimation of GNP). Profit is estimated as a residual (NNP - wages),
but this residual also includes the income of the self-employed (farmers,
artisans, carpenters, taxi drivers, etc.). So if in the process of the
development of capitalism, the relative proportion of this income of the
self-employed is declining, as is usually the case, then this appears in the
data as a decline of the rate of profit. Our group is trying hard to figure
out a way (at least a crude way) to separate out this income of the
self-employed, but it is not easy, and may not even be possible. Fellow
OPEL listmember Abelardo Marino and I have a meeting this week with
officials from the main agency for economic data in Mexico to see if they
can help us with these two main problems: the capital stock data and the
self-employed. The thought of trying to do this for a whole collection of
countries is mind-boggling (language barriers for one). But maybe if we can
do this for a small number of countries, we can will have a better idea of
the extent of the biases in the offical data. For the appropriate
estimation of rate of exploitation, there is the additional problem of
distinguishing between
productive and unproductive labor.
Comradely,
Fred