How was Homo Economicus manufactured? What's his shelf-life?

The design engineers in the manufacture of Homo Economicus.

1. David Hume
2. Adam Smith
3. Auguste Comte
4. Jeremy Bentham
5. James Mill
6. David Ricardo
7. John Stuart Mill
8. William Stanley Jevons
9. Leon Walras

Hume reminded everyone of the is-ought distinction (the difference manifested itself as political economy - Ricardo - versus moral science -  Malthus).

Smith centred economics on the understanding of human action - how we weigh choices and decide when engaged in economic behaviour.  That is, economic theory comes out of a general model of human choice-making.  In "The theory of moral sentiments", the self-interest motive was one of many - and certainly not the most important -  dimensions along which he considered man to operate.  Those other dimensions included other psychological facilities, and also the historically specific institutions which each economic agent finds himself in.  Smith's methodology amounted mostly to the advice - be empirical, be systematic; in particular he sometimes applied deductive arguments, sometimes inductive ones.

At the dawn of the nineteenth century, a kind of physics envy set in, resulting in an emphasis - by people like James Mill, David Ricardo, Jeremy Bentham - on the logico-deductive style of rhetoric and methodology.  At the same time, Auguste Comte was promoting a view of scientific methodology which categorised the sciences in relation to one another, in particular in relation of the more fundamental ones, the most fundamental of which was mathematics.  J. S. Mill cemented the concept of 'economic man' as a purely self-interested utility maximiser, no doubt helped by the preponderance of numbers and measures in economics (as opposed, say, to anthropology).  People like Ricardo Jevons, Walras contributed logical and mathematical rigour to the models of (self-interested) human action.  Then there was a move back out from an analyis of 'man in an economic setting' to a sociological view of man.   This movement was driven interestingly forward out of the domain of man-as-consumer - i.e. from the strict economic domain of interest - and into the human sciences generally by people like Wilfredo Pareto all the way to Gary Becker.  Here the key elements of the (neo-)classical approach:- 

1. a mathematocal or rigourous logical model
2. motivation reduced to pecuniary self-interest
3. a deductive reasoning style with an ahistorical flavour

James Mill and Ricardo were also responsible for promoting self-interest as the canonical human sentiment to use as the ideas-machine to generate their increasingly mathematical models.  We moved from a rich theory of moral sentiments of a single-minded soverign consumer in the space of a generation.  Ever since, economists have chosen where they stand with respect to how far the single-minded model gets you (rational expectations, efficient market theory, CAPM, Modigliani-Miller, Paul Samuelson versus John Maynard Keynes, Robert Shiller, Daniel Kahneman.)  The most amenable to numbers of all the human motivations is the one where we look to minimise costs and maximise our financial return.  It allows you easily to apply calculus and compounding to a tiny region of ideal human reasoning.  No wonder this became so popular a style of economic modelling - just like the drunk man looking for his keys under a lamppost since that's where the light is.  Actually, that analogy is perhaps too harsh.  Why blame this approach when it was clearly a decent avenue to pursue.  But something happened along the way - it changed from being a decent avenue to being the exclusive methodological approach to economic modelling.  And that is surely wrong.

J.S. Mill draws a distinction between a social statics and a social dynamics approach to human agency.  Comte, he claims, was strong on the latter, weak on the former.  It is the former which includes the domain of modelling which sees humans as financial maximisers, since this is generally considered to be more or less permanent (in the context of the last 13 thousand years of human history).

I've read Fiedman's famous "The methodology of positive economics" and I don't think it actually adds anything to the debate, which I consider essentially a philosophy of economics one.

Allen Oakley has a fantastically well written book, "Classical Economic Man", which is a history of the construction of the idea itself.  Specifically, he focuses his attention on Adam Smith and J.S. Mill and how the concept of economic man developed between these two great classical economists.  Likewise, Keynes's father wrote a book, quoted favourably at the start of the Friedman classic, which captures the methodological state of play as it stood at the end of the nineteenth century.

Bruce Mazlish also examines the idea of positivism in nineteenth century intellectual thought in a way that shows just how crucial this period was to all current economic work.

Jonathan Aldred and Justin Fox each provide a well reasoned modern critique of the idea and how new economic approaches are set to flourish post credit-crunch.

To say that the deductive model, with humans as exclusively self-interested - a mode eminently amenable to mathematical analysis - has been a productive direction is surely true.  It has opened up for us whole worlds, new instruments, new sciences, new industries.  It has surely transformed us too, as Donald MacKenzie clearly shows.  But is it now time to return to Smith, to a broader conception of economics based on a fuller range of human sentiments.  And surely we can find a way to mathematically, algorithmically, logically model these too?  Perhaps this will allow us to re-engage with the important institutions of economics, which have been denigrated and under-modelled for a long time.