GDP statistics - comment to David

Jurriaan Bendien bendien at
Wed Nov 12 14:37:22 MST 2003

Hi David,

Glad you found my remarks useful. In the 1980s I spent about 5 years working
on macro-economic statistics as research assistant and Phd student.
Subsequent to that, I worked nearly 4 years as research statistician in a
government statistics department. My activities directed me to what the
aggregates actually mean, how they fit together, and to what transactions
they refer, firstly as a socialist student making a disaggregate analysis of
statistical material in order to test Marxian hypotheses, and subsequently
in survey design, where I had to define and create statistical units and
aggregation principles in order to design official survey questionnaires and
standard classifications. You don't earn what an academic earns, you don't
have the status of an academic, the snobs give you no credit, but you learn
much more, and participate in work which has real consequences, beyond the
publication of an scholarly paper in a scholarly journal - the theorising is
disciplined by the necessity to collect and collate real data.

Often people associate statistics with probability theory and this is its
mystique, but in the real world we need to know exactly what statistical
units we are counting and how we count them, otherwise we manufacture "data"
which, although we can let plenty mathematical wizardry loose on those data,
do not mean anything, amalgamate different things, or mean something else
than they purport to represent. Karl Marx himself was vitally concerned with
categorical inquiry, but of course these days we can achieve much more
precision in this area. Towards the end of his life, Marx wanted to analyse
business fluctuations mathematically and statistically, but as Samuel Moore
commented to him, this wasn't really possible, because the data he needed
weren't (yet) available. Likewise Engels commented on the problem of
measuring value relations because of the paucity of data collection and
business secrets. Marx's categorisation and categorical inquiry was actually
so good, that he anticipated trends 50-100 years before they became an
observable commonplace.

If anything is endemic in the world of economics, it is woeful ignorance
about macroeconomic aggregates and the abuse of those aggregates (this gets
worse when postmodernist relativism, identity politics and subjectivism
begins to disorganise research), plus of course one statistic is
counterposed to another for ideological purposes, causing Disraeli to
complain about "lies, damn lies and statistics." One of the main reasons is
that the average economist does not look critically and systematically at
the social relationships behind the transactions, and confuses a statistical
summary description in thought with the real world to which it refers, but
beyond that, they just aren't familiar with the meaning of the concepts and
what is a valid indicator of what. I talked to an economics graduate once
who could not tell me what the components of GDP were or what it measured.
As I pointed out before, in reality neo-classical economic ideas are not
applied much in practice and cannot be applied much, but this follows
logically from the very ideology it propagates: market forces secure the
optimal allocation of resources, that is the dogma, and if that is so, there
is very little a conventional economist can actually do, except make
strategic comments on the bookkeeping (except when it concerns apologism for
misspending government and taxpayers funds). In which case, a socialist is
in some ways much better off doing a degree in accountancy or auditing,
which provides a far better insight into the core activities of economic
life in many respects.

I think you are correct that annual net output growth in the 1990s was on
average higher than in the 1980s in the developed countries. But this
doesn't really tell us very much yet about the variables and ratios Marx
specified as crucial for the "health" and survival of the capitalist system.
For Marx, an aggregate like GDP doesn't indicate fluctuations in economic
growth adequately at all, not just because you have to understand what GDP
contains, but because it abstracts from critical aspects of private capital
accumulation on which GDP growth ultimately depends, in a capitalist
society. In  Marxian analysis, what is much more critical as indicator, is
the fluctuations in real gross fixed capital formation within the
"producive" sector, i.e. the level of real investment in the production of
tangible goods and services, which is critical for overall employment
levels, and predicts the real dynamics of the system much better. Just
because real world GDP increases, this doesn't tell us very much about real
economic growth or the pattern of that growth, we have to make a specific
analysis of the relationships and transactions to which it refers. And in
fact Marx's own analysis applies precisely to that.

My own hypothesis is that the real production of the total mass of global
surplus-value (newly created surplus-value available for distribution) is
basically stagnant or (what I really think) declining, which creates a
tendency in the production process for a transition from the exploitation of
relative surplus-value to the exploitation of absolute surplus-value. As I
explained before, if the total economic cake is stagnant or declines, growth
can occur, only if more income from producers is transferred to investors
(but this growth does not necessarily in tangible production of net
additional material wealth, it could be just hotels and restaurants, air
transport, the FIRE sector and so on) but, beyond a certain point, you
simply cannot take more income off people because they haven't got it, and
so then they have to work more, and the resistance to the pressure to
increase the rate of absolute surplus-value is essentially a matter of
strength and combativity. The debt weapon is essentially a conduit for an
increase in the rate of absolute surplus-value.


PLEASE clip all extraneous text before replying to a message.

More information about the Marxism mailing list