Thursday, August 20, 2009

Keynes was right; then subsequently abused

It is a delight to be with you here today, discussing the policy prescriptions of the great John Maynard Keynes. (I like to pretend like it is an actual, face-to-face Ivy League-style debate.) Though I disagree with my opponent, I do appreciate his view. Unfortunately for him, it is just plain wrong.

Like all lazy speakers, I’m going to begin with a definition. Continuing from the wiki post Karl cited:
Keynes contended that aggregate demand for goods might be insufficient during economic downturns, leading to unnecessarily high unemployment and losses of potential output. Keynes argued that government policies could be used to increase aggregate demand, thus increasing economic activity and reducing unemployment.

Despite the insults that rabid anti-Keynesians hurl at the supposed evils of this idea, the theory is really quite simple, elegant and plausible. Economic output, as defined by GDP, is the sum of private consumption + investment + government spending + net exports or GDP = C + I + G + NX. When a severe economic downturn occurs, C & I contract (net exports are already a negative and have been for decades, but are also relatively small in comparison so we can ignore it for now). If consumption and investment are depressed, the government can, at will, grow its expenditures to prime the pumps of demand. This simple and mechanical relationship is the crux of Keynesianism. Is this so hard to understand? Apparently so. Let’s examine the arguments against this brilliant observation before turning to the supporting evidence.

Argument against #1: Government spending never ends; they become permanent entitlements and thus eternal debts.

Analysis: non-sequitur. Governments failing to use short-term stimulus says nothing about the validity of the theory itself. This only speaks to the politics.

Argument against #2: Deficit spending results in massive debt and inflation.

Analysis: Poppycock! Deficit spending has never resulted in inflation. The largest deficits relative to GDP in the post WWII era (present situation excluded) were during the Reagan years which coincided with disinflation! Monetary policy results in inflation.

Argument against #3: A dollar spent by the government does not have the same multiplier as a dollar spent by private enterprise.

Analysis: This argument refers to concept known as the money multiplier (MM). When Karl buys pop tarts from Canadian Tire for $1, CT must pay for the pop tarts, which in turn must pay for the sugar, wheat, flour, electricity, etc. The power and water companies must pay their workers and for the coal, etc. And so on. A dollar spent generates more economic activity than just the one dollar. We call this the money multiplier. Some like to assert that when the government spends money, it doesn’t generate the same amount of downstream activity because by its nature, government spending is inefficient and wasteful. To this I say “so what!” By its very nature, during an economic downturn, productive resources are being underutilized, which causes a lack of investment and then further slack demand. During such conditions, establishing demand for goods and services- no matter how wasteful or inefficient- employs otherwise unused, idle productive capacity. Fuller capacity means revenues, profits, a need to employ and invest in new capacity. A dollar is a dollar is a dollar. Period. Furthermore, at cyclical lows all multipliers are lower than normal precisely because of unused capacity, so it is no wonder that government stimulus spending is thought to have a low multiplier.

Argument against #4: No one can calculate the amount of spending and duration needed in order to achieve full employment.

Analysis: Straw-man. The theory, as was always taught to me, is a counter-cyclical tactic. If it is misused or applied to that which it wasn’t intended, that isn’t Keynes’ mistake or a fault in the theory. Using it in an attempt to remove the business cycle is ridiculous.

Argument against #5: If Keynesianism is so great, why not increase government expenditures to 40% or 50% or 75% or 100% of the economy? That should work itself out brilliantly, you socialist, communist, fascist, moron! (or whatever you like to call me).

Analysis: In logic and many thought experiments, it often helps to examine the same situation from extremes to check validity. However, Keynesianism does not reject capitalism as a means to allocate resources as name callers would suggest. It is another straw man to say Keynesians want to destroy capitalism and replace it with government. Nothing could be further from the truth. It simply aims to boost overall economic activity while the economy rights itself from within.

Consider that an economy is a system. It is a system which allocates resources by balancing supply and demand. Occasionally, the systemic stabilizers can get so far out of whack that it doesn’t right itself without creating immense pain and suffering. Will it ultimately fix itself? Probably, if the people don’t revolt first. Which is better, insist on letting it fix itself and see if the country can hold together or try to help it along? An analogy might be a high fever. Sure, we all get them and usually survive after some pain and suffering. But then there are those pathogens that overwhelm the system and the natural mechanism- fever- that kills the pathogen. The fever gets so high and persistent, it begins to kill the host with - you -with dehydration, delirium and other symptoms! When such a vicious cycle of deterioration has begun, the answer is medicine. To hell with letting the body fix itself, give me the meds!

Before we take a look at the evidence suggesting Keynesian economic policy works, please remember that macro economics can’t be tested in the lab. The few times Keynesian policies were enacted in earnest, other things were going on too. So I’m going to concentrate on the biggest and baddest of the Keynesian experiments (and no, I’m not going to present Roosevelt’s New Deal as evidence; Too many bad policies worked against the good Keynesian ones to make any claims.) I can think of two solid examples of Keynesianism at work: WWII and the militarization of Nazi Germany. The latter is a better example and offers clear evidence that Keynesian spending can break the vicious cycle of depression. There is nothing less efficient and more wasteful than military spending and both ended the depression.

As for the assertion that Keynesianism is bunk simply because it spawned Neo-Keynesians, New-Keynesians, etc.- well, we started with gravity, progressed to relativity and then special relativity. We began with a simple atomic model of electrons, protons and neutrons. We ended up with quantum physics and now possible string theory. First came the Magna Carta; later came the U.S. Constitution. Theories get refined- especially the good ones. Need I say more?

10 comments:

  1. For argument #2, you leave out the question of debt. Basically, Keynesian stimulus is a way of borrowing from the future to consume in the present; a way of smoothing total consumption.

    However, paying back the debt in the future (with taxes, mostly) could have a negative effect larger than the positive effect stimulus has, which makes it a wash as fiscal policy.

    Increasing debt now only has a nasty effect on borrowing later. A large stimulus now, in a W-style recession, could make it harder to borrow in the second trough, which may be deeper. While this is an argument against the application of a Keynesian stimulus, more than the theory itself, it does call into question the idea that governments can borrow as much as they need to fuel their stimulus. If they can't, a proper Keynesian stimulus cannot be applied.

    ReplyDelete
  2. On this thought that 'stimulus now borrows from future growth', I am not so sure. On the surface, it seems to make logical sense. However, I can think of 3 reasons it may be otherwise:
    1) Doesn't current investment lead to greater future productivity? Make the roads better now and it reduces the waste that is commute traffic for years to come; improve education and [obvious]; improve energy efficiency and you reduce future costs while reducing dependence on volatile inputs; update computer systems and get more efficiency later. (I realize many believe in the gov multiplier being smaller than 1, but that is tangential to this argument.)
    2) Simple math: When you compound growth rates, a less volatile series results in greater absolute growth. For example, if we have 3%, 5% and 7% growth in consecutive years, this 5% average growth results in 15.7% total growth after the 3rd year. But if this series is instead 5%, -1% and 11%- a very volatile series- it results in 15.4% total growth while the average is an identical 5%. Maybe smoothing isn't so bad after all?
    3) Creative destruction seems an obvious argument against #2. Many of the companies that would go bust in a more extreme environment SHOULD go bust. We don't want to keep them afloat. This is true, but won't they go bust anyway? It may take an extra year, but they aren't competitive anyway. Plus, dismantling too much economic infrastructure could be disruptive and anti-efficient (so to speak) in other ways. Bernanke argues in his papers that the financial intermediary function of banks and the stock market was so disrupted and impaired during the Depression that it actually delayed recovery by preventing investment.

    Anyway, I'm not wholly convinced in the smoothing hypothesis just presented. I do think it needs a more thorough vetting though. Particularly #1.

    ReplyDelete
  3. Brett, I really enjoyed your defense, particularly when you trotted the Nazis out as a good followers of Keynes. LOL

    Some thoughts on the last post.

    1) Current investment doesn't necessarily lead to future productivity. The better question is what provides a better return on investment: government spending or private spending. The answer is private spending, so government spending should be held to the absolute minimum of services and infrastructure that the private sector can't or won't provide. You are also assuming that government can improve on the job the private sector does when it has no track record of improving it's own efficiency or productivity, maybe with the exception of the military.

    2) When have we had GDP growth that either annually or quarterly had 5%, -1%, and 11%?

    3) Companies are self interested in prolonging their pain so they will do whatever they have to in order to survive. The question is whether the government should have a role in keeping them on life support. The sooner you hit bottom, the sooner you can resume real growth.

    ReplyDelete
  4. Argument against #6: Keynes took the effect and made it the cause.

    The crux to Keynesianism is that the overall desire to consume falls well short of the overall production capacity. The false premise is that the problem to economic life is the production of consumption. It wrongly assumes the problem of production has been solved.

    In fact the opposite is true. The fundamental problem to economic life has always been and always will be the challenge to produce or how to expand production. Demand is insatiable, however consumption is limited by the ability to produce. An individual must first produce something of value prior to exchanging it for something she desires to consume.

    If one consumes a product or service without offering like value in exchange then one consumes at their expense. In this case the federal government consumes goods and services at the expense of the taxpayer or future debt. Or if the treasury prints new money and then uses it to purchase products and services without producing equivalent value then someone must suffer a corresponding loss. This loss is comprised of a decrease in consumption, depletion of capital, or a lack of reward from labor (i.e. higher taxes). The total of that loss equates to the same amount of products and services obtained by the government who did not produce.

    Only individuals can produce wealth. Governments do not, but nearly all of them are experts at confiscating it.

    ReplyDelete
  5. 1) Infrastructure type spending is what Keynes had in mind. Not Cash for Clunkers or floating GM. Infrastructure, whether that is roads, hydroelectric or education is productive. Clean water is productive and a public good not likely to be achieved without government. How is a better road or education not productive?

    Where do I "assume the government can improve on the job the private sector does"? That is not an assumption- government spending is only to replace a portion of insufficient aggregate demand during a severe downturn.

    2) Those numbers were for mathematical illustration pulled from the air. But since you ask:
    1932-38: -13%, -1, 11, 9, 13, 5, -3
    1948-50: 4.4, -.5, 8.7
    1953-55: 4.6, -.7, 7.1

    3) "The sooner you hit bottom, the sooner you can resume real growth."
    That is the crux of the social argument here: better to ease the pain and possibly push out some faster growth or suffer a lot more and risk the social unrest that comes with it in order to resume fast growth sooner? You are arguing that deeper cycles (ups and downs) is better than smoother cycles in all aspects. I'm saying there is a social cost to the great swings that deserves consideration.

    ReplyDelete
  6. Your comment is right on the money and I mostly agree. However, I disagree with your premise that "The crux to Keynesianism is that the overall desire to consume falls well short of the overall production capacity."

    In a depression, it is not the desire to consume that falls well short, rather it is the ability to consume. Keynesianism directs the gov to consume, because it is the only that can.

    ReplyDelete
  7. OK, never ask a question you don't know the answer to. Rookie mistake. So I went back and looked at the data going back to the beginning of the 19th century and there have always been pretty substantial swings in GDP that seem unaltered by Keynesian spending. I hear the argument "think how bad it could have been" but how do you prove it?

    ReplyDelete
  8. Debt rarely gets paid back. We simply pay interest on it forever as the debt is continually rolled over. If an economy can recover and grow faster than future debt, then the debt burden shrinks in relative terms.

    To paraphrase you: governments failing to pay debt back says nothing about the validity of my theory itself. This only speaks to the politics.

    If we have to debate this, we need to stick to one realm or the other.

    ReplyDelete
  9. I have some criticisms of Kenynsian economics, and I'm interested to know what you think of them:

    1.) Time lags. Monetarists claim that fiscal policies have such a large time lag that by the time they take effect the economy may have continued to do so poorly that the policy is now moot. Therefore, governments should compliment (or even replace) fiscal policies with monetary policies - even if just in the short run. Do you think they may have something there?

    2.) What if people are forward looking and realize that government spending will one day be paid for with a tax increase? It may be that then people will reduce consumption in the present(this effects investment as well) because they know its coming. Thus, the effects wash out. [This is basically the Rational Expectations Hypothesis from the New Classical school]

    3.) If the government doesn't use taxes, they will have to borrow - and this may exacerbate the problem of the Twin Deficits. (http://en.wikipedia.org/wiki/Twin_deficits#Proof)

    --BTW, this is my first time finding this blog - I saw it mentioned over at skepchick and I'm really pleased to see a quality critical thinking economics/finance blog. I think this is an area that critical thinking has miserably failed at getting into thus far. So thanks, keep up the good work!

    ReplyDelete
  10. dp- Welcome to the blog and thanks for the great questions. Let's take a closer look:

    1) Time lags, indeed. Yet in the depths of a depression, 3, 6 or 9 month lags are irrelevant. Breaking the vicious downward spiral is almost as much sentiment as actual economics when unemployment is 15 or 20%+. Also remember that tax cuts are almost immediate, so the time lag can be reduced. Re monetary policy- since the Fed is quasi-governmental and controls monetary policy, this form of stimulus is already used widely and primarily.

    2) Alternatively, why wouldn't people consume today while they have the money instead of postponing it to when they don't have the money because of taxes? Use it or lose it seems more rational to me. And since when do people rationally expect anything? When they took out liar loans to buy wildly inflated homes? When they paid infinite P/Es to get the latest dot com? When they lend trillions to people to buy inflated homes? When they buy stocks at the top (as fund inflows indicate they always do)? When they pay 20% interest on credit cards to buy something now? Rational expectation-ists need to review some behavioral finance findings.

    3) Twin deficits are indeed a problem. No argument there. However, in a depression which is more harmful, allowing the economic death spiral to continue or incurring debt to break the cycle? We often borrow to get ourselves out of a pinch. If it isn't creating a perpetual state of borrowing, then "one-time" debt shouldn't be a problem. Alternatively, in another context, say that increased debt causes the currency to decline. A weaker currency causes lessened ability to import, which is the cause of the current account deficit. The weak currency also boosts exports. Thus it is possible for the extra debt to help speed the correction of over-consumption of imports.

    ReplyDelete