In neoclassical DSGE models, coastal defences are ‘wasteful’. A reply to June Sekera

Dear mrs. Sekera,

thank you for your very clear and insightful blogpost on this blog about economists and their distorted concept of public goods. But I’m afraid you’re way too positive about mainstream macro models and ‘public goods’. Often – these models do not have a concept of public goods.

On a regular basis neoclassical ‘macro’ models do not have any logical space for public goods or ‘government consumption’ (i.e. consumption by households produced or financed by the government, like education). Which is, of course, daft. My country – the Netherlands – would not even exist without coastal defenses. And believe me – we learned about the necessity of well maintained large-scale public coastal defenses the New Orleans hard way (Knibbe, forthcoming). After every major flood there was a wave of public centralization and extension of planning, design, control and financing. ‘Wie niet dijken wil moet wijken’ (those unwilling to pay and build a coastal levee – must flee). We actually built a statue for Caspar de Robles, a Spanish oppressor who was kicked out during the Dutch Revolt, because he improved financing and building of levees in Friesland in the sixteenth century. But in the default neoclassical macro economic DSGE (Dynamic Stochastic General Equilibrium) models, public expenditure on coastal defenses is, by definition, ‘wasteful’ – as it’s government expenditure.  And government expenditure, all of it, is wasteful in these models as it uses resources which could be used for private consumption. A nurse paid by the UK National Health Service is ‘wasted money’,  a privately paid nurse adds to utility. That’s the state of mainstream macro-economics, neoclassical style. You are attacking Samuelson – but Samuelson crafted his ideas (which as you state do have basic flaws) to combat exactly this kind of thinking! He clearly failed. This kind of thinking is alive and kicking. And kicking hard – as the over five million unemployed in Spain can testify. Even incorporating the flawed ideas of Samuelson into these models would be a huge step forward. These at least consider the existence of public goods and services.

The Oosterscheldekering coastal defense. Wasteful?

There are endeavors to change this. But the very fact that these articles have to state, in an explicit way, their difference with ‘standard’ neoclassical ‘macro’ by introducing the notion that public expenditure can serve a purpose shows the ‘state of the mainstream art’: unmitigated market fundamentalism. Don’t misunderstand me. I do consider governments to be dangerous entities as, in Europe, the shift of power to undemocratic institutions like the ECB and the Troika shows. Also ‘the left’ should embrace all efforts to attain a world guided by less and simpler rules. A goal which might only be attainable by limiting the number of government bureaucrats. Despite this caveat I like ‘my’ coastal defences very much (financed and designed by the government, built by private companies). Also, the teachers of my children do an excellent job. Dutch roads are excellent. Such public goods and consumption are, however, neglected by DSGE-economist or, to phrase it in their language, ‘public consumption does not enter the utility function of the representative consumer’. Not even in a Samuelsonian way. Remarkably, the neglect of government products and services is not ordained by the logical structure of the models. It is a choice by the economists designing these models, as models which do admit the existence of government products and services show.

Two examples:

The first example is an article by Yasuhara Iwate (2012) aptly titled: Non-Wasteful Government Spending in an Estimated Open Economy DSGE Model:Two Fiscal Policy Puzzles Revisited.

The naive among us might read this title the wrong way, as it seems to implicate that DSGE models as a rule make a division between wasteful and non-wasteful government spending. But it doesn’t. The title is meant to indicate that the article departs from the general habit of these models of treating all government spending (coastal defenses, education, health) as wasteful: ‘very few papers have considered non-wasteful nature of government spending‘. It does this by introducing the notion of ‘Edgeworth complementarity‘, i.e. the ‘revolutionary’ idea that people privately owning cars benefit from the existence of (public) roads. This is necessary to solve two supposed puzzles: “Contrary to theoretical predictions, recent empirical evidence suggests a crowding-in of consumption and a depreciation of the real exchange rate after a government spending increase“. Wow. More and better roads and bridges leads to more traveling. To economists, this is a genuine puzzle. Households (the household…) in these models are ‘rational’ and ‘Ricardian’ in the sense of the economist, i.e. increasing wages of teachers is supposed to lower consumption this wage increase is wasteful… Which is wrong when we look at the behavior of real households. But the puzzle is solved when we consider the fact that government consumption might have a useful purpose, like high quality education.

The Millau bridge – public/private partnership. Wasteful?

The second example is a 2011 article by Stähler and Thomas with titled: “Fimod – a DSGE model for fiscal policy simulation“.

This article takes the ‘revolutionary’ approach that parents actually like it when their children get an education: public consumption directly enters the utility function of the representative consumer. There are actually two consumers in this model but they do not interact in a genuine social way as, contrary to real life, their interactions do not change their attitude, their sense of self or their balance sheets. Which means that ‘utility’, as used in this model, still adheres to the neoclassical concept. A fun part of the article is that is uses the concept of the rational Ricardian household which increases expenditure (as long as it can borrow) when the government cuts wages and dismisses people as the household perceives this as ‘less waste’ and hence more space for private consumption:

Because of lower private wages, less public employment and an increase in unemployment, consumption of RoT households [households which can’t borrow, M.K.] falls, while it increases for optimizing households due to anticipation of lower future taxation; the latter effect dominates, and thus total private consumption rises.

This is called expansionary austerity and it does not exist. Cutting government wages and sacking government employees did not exactly lead to a consumption surge in southern Europe.

But that’s all about private consumption, back to public consumption. What happens when we concede (as this model does) that government financed education is not wasteful by definition (i.e.: ‘enters the utility function’)? The model becomes kind of indeterminate in its results. The joyful austerity idea that a crisis is best solved by crushing the government vanishes into thin air, even when direct beneficial effects of government consumption are excluded:

However … in our model, we neglect any potentially positive (say, for instance, private productivity-enhancing) spillover from public employment to the private sector. Government services only enter households’ utility [i.e. not the ‘production function’, learning children to read and write has no productive use, M.K.]. As it is not clear how much public employment (positively) affects private sector production, we chose to abstract from this issue. Nevertheless, the positive effects of public sector employment cuts found here are certainly diminished in practice (or may even vanish when public employment reduction is misplaced) … we find a cut in public sector wages to be beneficial for private sector output and for both private sector and total employment. … Our simulations suggest that cutting public sector wages is, in the short and long-run, the most efficient consolidation strategy in terms of economic activity, i.e. production and employment … However, again, we must point to some caveats regarding this finding. Cutting public sector wages (too much) may imply that the public sector no longer finds (qualified enough) workers to do the relevant tasks, or that public sector workers would no longer provide the necessary effort (a possibility the literature on efficiency wages points to). In both cases, the provision of public services may be affected. Given that public services enter the households’ utility, such measures may not necessarily be welfare-enhancing even though they increase economic activity. Furthermore, our model neglects the possibility that government services indirectly foster private-sector productivity.

Dear mrs. Sekera – these articles are the exception. They concede a possible, ‘Samuelsonian’, positive role for ‘public goods’. But the ‘default setting’ of these models is the idea that public expenditure is wasteful, by ideological nature. That’s a lot worse than Samuelson whose ideas, despite all their shortcomings, were in crafted to combat that kind of nonsense! We have to continue this struggle (at this moment in the Euro Area by enhancing democratic control of central banks). Samuelson forged the wrong weapons but even his ideas are not embedded in mainstream macro models, anymore.

Again, the government is a dangerous entity. On the other side – the heroic work of Savitribai Phule (1831-1897), the first women’s educator of India, is far from completed. And we need governments to do this. To much talent is still wasted, to many women – and men – are not yet empowered by public education. Economic models should admit this.