Welcome to Quadrant Online | Login/ Register Cart (0) $0 View Cart
September 01st 2010 print

Peter Smith

Time to Topple Keynesian Economics

When we look at profligate governments, those in debt and getting deeper into debt, like those responsible for governing the UK, the USA, Greece, Portugal, Spain, France, Ireland and others; there are presumably professional economists in the background. Over the past ten to twenty years of government overspending, some senior public servants, trained in the rigorous discipline of economics, must surely have been giving consistent and fearless advice that it would all end in tears. Or have they?

Take the massive government spending employed by most countries to combat the global financial crisis (GFC). There was no evident sign that senior public sector economists were urging some modicum of restraint. The reverse appeared to be the case. Ken Henry was certainly not urging restraint. We need a stimulus package and we need it now, summed up his position. The US Treasury Secretary, Timothy Geithner, was an enthusiastic advocate of stimulus spending and apparently remains wedded to still more, if his comments at the G20 meeting in Toronto in June are any guide.

Wherever we look, the economics profession in government service appears to have been hijacked by interventionist Keynesian economists. Politics is part of this but it is by no means the whole. Some so-called economists are left-wing ideologues. Most, however, are simply victims of someone described colourfully by Thomas Woods in Meltdown as “one of the twentieth century’s crackpots”: John Maynard Keynes.

Crackpot or not, Keynes has spawned generations of economists and politicians who think that spending money is the same thing as making money. You might recall Michael Douglas’s character in War of the Roses saying to his wife (played by Kathleen Turner): “It’s a lot easier to spend it than it is to make it, honeybun!” Well, not according Keynesian economists, it isn’t.

Keynesian economists believe in the spurious and curious notion that economies are driven by spending rather than by making. To them spending is an economy’s driving force. There are three mutually reinforcing factors at work that promote and sustain this belief. One is the way economics is taught; a second is the formation of a “conventional wisdom” around Keynesian economics; the third is the way the national accounts and macroeconomic modelling “bulletproof” this conventional wisdom against any and every experience. I will go through these factors in turn before attempting to assess whether the times make Keynesian economics ripe for toppling.

Economics teaching

For the most part, economic advisers to governments have been taught macroeconomics through the prism of Keynesian economics, as though they are one and the same. They come away believing that boosting aggregate demand through government spending is the way to produce growth and prosperity.

The key to this is the word macroeconomics itself. It has a clouded but relatively recent history. The edition of the OED that I consulted listed its first use as 1948. It seems likely however, based on a number of sources, that it was first used by the Norwegian economist Ragnur Frish in 1933. But it was not in common use, nor was its counterpart, microeconomics, until the second half of the twentieth century. Importantly, the term only came to prominence in a Keynesian context. Keynes did not use the term in the General Theory but he, nevertheless, is its true father. Keynesianism and macroeconomics became synonymous, as they still are to an overwhelming extent.

Classical economists of the eighteenth and nineteenth centuries, from Adam Smith to J.S. Mill (and neoclassical economists like Marshall and Pigou who followed them in the late nineteenth and the early twentieth century prior to Keynes) conflated the big picture—the nation and trade—with the small picture—households, workers and firms. They had it right because the big and small pictures are inextricably linked. You can’t look at one without considering the other.

Keynesianism produced a pedagogical schism in economics that has no relevance at all for the way economies work. Textbooks were divided into two parts: microeconomics and macroeconomics. Macroeconomics took on a life of its own. It was and still is a Keynesian world of aggregates, seemingly distinct from the economic forces at work among households, workers and firms. Keynes and his followers invented a new artificial world which, in fact, has no relationship to underlying economic forces. It has no underpinning. It has no legs. Trying to understand how the economy works, and how to put it right, through macroeconomics is like trying to understand how an engine works, and how to fix it, without regard at all to its constituent parts and how they work together.

This is why public sector economists and most academic economists and economic commentators are currently floundering, along with the economies in receipt of their Keynesian prescriptions. They wonder: What has happened? We have spent big, borrowings are astronomical, but our economies are still in recession and getting worse. They have no unified theory of economics to look to for guidance.

Learning economics since Keynes means having to develop something akin to a split mind. You must learn all about microeconomics and the complex way market economies work. You then must put that totally aside and, preferably, out of mind while you learn macroeconomics, which pays not the least heed to the complex way economies work.

Paul Krugman, writing in the New York Times after the June G20 meeting, despaired that another depression was in the offing. He lamented that governments were “preaching the need for belt-tightening when the real problem is inadequate spending”. At least you can say of Paul Krugman that he is consistent. He believes in a fiction to the bitter end. But what of those economists previously preaching from Keynesian texts, who are now advising governments to tighten their belts? What are they now using for a text? The fact is they don’t have one. There is one—it can be labelled pre-Keynesian economics—but we can safely assume they know nothing about it. They are flying blind and we are all passengers. They’re not letting on because we might panic and, perhaps more importantly to them, they would lose their credibility.

The question that obviously springs to mind is whether the evident failure of Keynesian prescriptions in the wake of the GFC will undo Keynesianism and the unwitting charlatans that practise its black arts.

Conventional wisdom

A good question is best answered by a definitive answer. I don’t know that I have a definitive answer. It is clear, however, that Keynesianism will be hard to dislodge, since it is a powerful “conventional wisdom” of the kind identified by J.K. Galbraith in 1958 in The Affluent Society. And, more than this; it is a conventional wisdom that has surrounded itself with a defence mechanism to protect it against what Galbraith called “the march of events”.

Galbraith explained that once established, conventional wisdoms are tenacious. They resist challenges vigorously and without compunction.

They have a large literature, even a mystique. The defenders are able to say that the challengers of the conventional wisdom have not mastered their intricacies … The conventional wisdom, having been made more or less identical with sound scholarship, is virtually impregnable. The sceptic is disqualified … were he a sound scholar he would remain with the conventional wisdom.

Only the persistent inconsistency of the march of events, Galbraith explained, could eventually undo a conventional wisdom: “while the world moves on, the conventional wisdom is always in danger of obsolescence. This is not immediately fatal. The fatal blow … comes when the conventional ideas fail signally to deal with some contingency to which obsolescence has made them palpably inapplicable.”

Certainly, the malaise which has followed the Keynesian response to the GFC is a march of events which might objectively be viewed as fatal to Keynesianism. Government spending has singularly failed to restore economies to health and has left governments wallowing in debt. But that would underestimate the defence mechanism that this conventional wisdom has at its disposal. This defence mechanism is in the form of national accounts and macroeconomic modelling. It works, or up until now it has worked, to bulletproof Keynesianism from the march of events.


Modern national accounts mirror Keynesian theory. Demand in the form of expenditure aggregates (such as residential and non-residential construction spending, consumer spending, and government spending) is on one side; production is on the other. Any increase in a particular expenditure aggregate is shown dollar for dollar as contributing to production (GDP).

Most analyses of national accounts appear to lack any real insight as to how they are put together and their meaning. Politicians usually have a story to spin. They want to show that their actions, including any profligate spending, have had good effect. But, economists should display a better appreciation of the numbers and provide a balanced view. That simply doesn’t happen.

Economists in and outside public service are prone to present the national accounts at face value and report that Keynesian stimulus spending has worked. The accounts provide no such evidence. It is a mirage. If government spends a dollar; this will always be shown as contributing to GDP in the period it is spent. And, it would be shown this way even if the whole dollar were spent on imports and did not, in fact, contribute to GDP.

Money flowing through the economic system is fungible (you can’t tell one dollar from another). So, was that particular dollar spent on imports or home-produced goods? There is no way of telling. If the dollar were spent entirely on imports, the national accounts would show a contribution from government expenditure and an offsetting negative contribution from imports. There would be no way of determining that the dollar in question was the one spent on imports and, even if there were, the accounts would still be presented as they are.

The fact that government expenditure is shown dollar for dollar as contributing to economic growth is therefore meaningless. It is a truism because of the way the accounts are constructed. It doesn’t show that stimulus spending works at all; it simply reflects the way the national accounts are put together.

If the government spends taxpayers’ money to build, say, a school library, at a reasonable cost, this will be shown dollar for dollar as contributing to GDP. If the same library were built at an inflated cost, this would be shown dollar for dollar as contributing to GDP. If the same library were then immediately demolished at government expense even more would be added to GDP. And this is not the end of the matter. If government takes resources away from the private sector to build a school library, this will be recorded, and probably written up, as the economy being saved from a decline in private sector activity by government spending.

The post-GFC world has been one of economic stress when economics matters more than it does in more stable economic times. The penalties for wrong-headed economic thinking are potentially more onerous. Constructing, and giving currency to, a baseless interpretation of national account figures is not helpful. A rudimentary understanding of the national accounts, as taught in any basic economics course, is all that is required to interpret the numbers properly and bring some clarity to the economic debate.

Unfortunately, the national accounts are only part one of the obscuration of the economic debate. Clarity suffers still more, and unwarranted credibility is given to Keynesian policies, when public sector economists turn to their macroeconomic models.

However complex the model, it will be based on the textbook macroeconomic identity that aggregate private consumption spending plus private investment spending plus government spending plus exports minus imports equals production. It will also have within it a positive relationship between an increase in production and an increase in employment.

If you impose on these models an increase in government spending they will inevitably show an increase in production and employment. It is important to understand that whether total employment rises or falls, the models will still show that an increase in government expenditure will have contributed to employment growth. Because of this, in the face of widely different economic experiences, both the Australian government and United States administration were able to claim that stimulus spending had generated many jobs.

As part of this year’s budget papers, the Australian government claimed that stimulus spending had created or saved 225,000 jobs. The United States administration claimed in mid-July that its stimulus package, rushed into law on February 17, 2009, for “shovel ready” jobs, had created or saved 3 million jobs. These jobs in Australia and in the United States were not physically surveyed and counted, and assessed as to whether they represented net gains. They fell out of macroeconomic models. The fact that overall employment had risen in Australia since the beginning of 2009 and fallen in the United States by over 2.3 million jobs (from the end of February 2009 until the end of June 2010) was of no apparent consequence. Employment rises; employment falls; it doesn’t matter, the stimulus has worked. At least Christina Romer, chairing the White House Council of Economic Affairs, when announcing the unbelievable gain of 3 million jobs, had the grace, to my eyes at least, to look embarrassed.

In fact, there is no basis at all for thinking that stimulus spending created the number of jobs claimed by either the Australian government or the United States administration. They are fictions which have gained credibility because they have been put on the public record. The models cannot be used to show that a specific amount of government expenditure will create or has created a specific number of jobs.

When I first started learning about statistics I was presented, as most students of statistics are, with an intended humorous example of the pitfalls of drawing conclusions based on correlation. The example I was given was the apparent relationship between the rising number of ordinations of priests and the rising crime rate on the streets. The example illustrated the potential silliness of mixing correlation with causation. Yet all statistical models, however complex they are, must still rely on time-series or cross-sectional data and the measured correspondence of particular variables, one to the other, to point to the possibility of causation. There is no getting away from that. Either we know absolutely that one thing causes another (heat applied to water in sufficient quantity causes water to boil) in which case we don’t need statistical models, or we are dealing with unknown causal factors and are relying on statistical modelling of apparent relationships to provide estimated formulas of what might have caused what and by how much.

Social scientists are well aware of the pitfalls of modelling and the risk of reaching false conclusions. Two variables can appear to be causally related and yet, in fact, be both related to a third factor and be simply fellow travellers responding in lockstep to that third factor. In that case, the causal relation may appear to be robust when it is in fact non-existent. Out-of-sample estimation is particularly fraught. There is no way of knowing that a relationship measured over data stretching into the past will have any explanatory power in the current period or in the future. For example, there is no doubt that real GDP and employment in Australia over, say, the last ten to twenty years, have grown in close tandem as have real GDP and government expenditure. However, it would be invalid to conclude from this that a specific increase in government expenditure will result in a specific rise in employment. Economic life is more complicated than that, and yet this conclusion is precisely what Keynesians ask us to swallow. If it is swallowed, Keynesianism survives any march of events: if total employment falls then Keynesianism has saved jobs and prevented things from being worse; if overall employment rises, Keynesianism has created jobs and saved the economy from recession.

I am not just saying that macroeconomic models are incomplete, though that is true. They cannot capture or model all of the negative and disruptive effects of increases in government expenditure on private sector activity. What I am saying is that macroeconomic models are unfit for the purpose for which they are employed. They will always provide the same answer no matter the circumstances. They are built to provide an answer in keeping with Keynesian policy, while all the time being held up as providing useful information on the success or failure of that policy.

At question is whether, despite the bulletproofing, Keynesianism will succumb to the post-GFC march of events in Europe and in the USA. To use Galbraith’s terminology, is the “contingency” of recession and debt, following massive stimulus spending, of sufficient weight to show that Keynesianism is “palpably inapplicable and obsolete”. To repeat the question I posed before: will Keynesianism be undone?

Undoing Keynesianism

The short answer to the question is that I don’t know but it should be.

Keynesians believe that economies are driven by demand. If private sector demand is insufficient they believe that it must be supplemented by government demand to ensure full employment. The key to undoing Keynesianism—if it is to be undone—is to hold it to account.

When resources are unemployed, Keynesians believe that government spending will always have net beneficial economic consequences. Government spending will result directly or indirectly in sufficient real economic gain to more than offset any attendant real economic costs. There may be costs. Keynes himself raised the possibility of government spending and borrowing crowding out some private sector demand. Nevertheless, any such costs are assumed to be less than the gain.

If Keynesians do not believe this; then I don’t know what they believe. I get a good sense that Paul Krugman believes it and perhaps Geithner. But what about all those governments and their economic advisers who are now preaching spending restraint when their economies are recessed; with plenty of unemployed resources ready to be put to work? Are they all Keynesians still? What do they believe now?

If Keynesianism works it works. Keynesians should explain with conviction, despite massive government debt, that it would be beneficial for European economies to stimulate more. They should explain that this would produce net economic gain and a better ability to service debt. If they were convincing enough, perhaps capital markets would be supportive. None of that will happen, of course, because most of them have lost their Keynesian faith; though they probably still cling to its trappings.

The time is right to lose the trappings too. That requires something to be put in its place. That something is an appreciation that the economy has to be seen as economists before Keynes saw it; as a whole and complex collection of many interrelated parts, not as a limited set of amorphous aggregates.

The European and United States debt crisis is a march of events to which Keynesianism has no answer. It is palpably failing and should be discarded. The times are demonstrating its poverty. It was always poverty-stricken and enervating but its deleterious impact was hidden by economists wielding meaningless and deceptive proofs of its success; by the resilience of market economies in the face of government economic mismanagement; and by the fact that even profligate governments take time to run down rich estates.

There is recognition within economics that governments provide enormous and essential assistance to economic progress by enshrining property rights within the law. Government is seen as providing a legal umbrella within which markets play out. Government economic policy should be similarly nuanced. The guiding principle should be to provide a stable, supportive and flexible environment within which market economies, and their constituent households, workers and firms, can prosper in normal times and, importantly, adjust in distressed times.

Politics and economics appear to go hand in hand these days. Economics should be objective. Economics is subjective in the hands of the Left who particularly favour the interventionist character of Keynesianism. It is then not a question of whether it is right but whether it satisfies a worldview that capitalism needed saving from itself by big government. The debt crises in Europe and in the United States make it evident that big government is the problem and that Keynesianism is part of that problem.

Economists, particular those in government service, should consign their existing macroeconomics to the dustbin and, perhaps, start reading some older texts. They certainly need to build their macroeconomics around microeconomic foundations, not as a disembodied artefact. As hard as that all is, unless it is done, toppling Keynesianism might be like toppling socialism. It will keep popping back up. 

Peter Smith is a frequent contributor to Quadrant and Quadrant Online. 

Peter Smith, a frequent Quadrant Online contributor, is the author of Bad Economics