Wednesday, June 6, 2012

Cutting Interest Rates to Boost the Economy: A Fallacy of Composition just like the False Notion of Cutting Wages to Create Jobs?

Does monetary policy work? Can the central bank meaningfully impact the economy and employment by altering interest rates? Or, as Warren Mosler sometimes suggests, are central bankers like the kid in the back seat turning the plastic steering wheel and thinking they are controlling the car?

Either way, most mainstream economists are firm believers in the power of monetary policy as a tool for steering the economy.

Let's first consider a different example of economy-wide price changes. Here is Paul Krugman (a prominent mainstream economist) on the fallacy of composition regarding wage cuts leading to greater overall employment (color emphasis mine):
"So let me repeat a point I made a number of times back when the usual suspects were declaring that FDR prolonged the Depression by raising wages: the belief that lower wages would raise overall employment rests on a fallacy of composition. In reality, reducing wages would at best do nothing for employment; more likely it would actually be contractionary.

"Here’s how the fallacy works: if some subset of the work force accepts lower wages, it can gain jobs. If workers in the widget industry take a pay cut, this will lead to lower prices of widgets relative to other things, so people will buy more widgets, hence more employment.

"But if everyone takes a pay cut, that logic no longer applies. The only way a general cut in wages can increase employment is if it leads people to buy more across the board. And why should it do that?"
(To see an illustration of why wage cuts won't create jobs overall, try the EconViz pages on how spending equals income at the macroeconomic level and about price level changes aka inflation.)

Just for fun, let's change the word "wages" to "interest rates" for two sentences from the quote above, and inject one example ("houses"):
"...the belief that lower interest rates would raise overall employment rests on a fallacy of composition. In reality, reducing interest rates would at best do nothing for employment; more likely it would actually be contractionary... The only way a general cut in interest rates can increase employment is if it leads people to buy more [e.g., houses] across the board. And why should it do that?"
Word games do not a proof make, but let's work through how this fallacy of composition might apply in the case of interest rates as well as wages.

So, what is the transmission mechanism by which interest rate changes supposedly help steer the economy? Paul Krugman again (emphasis mine):
"Back in the old days, when dinosaurs roamed the earth and students still learned Keynesian economics, we used to hear a lot about the monetary “transmission mechanism” — how the Fed actually got traction on the real economy. Both the phrase and the subject have gone out of fashion — but it’s still an important issue, and arguably now more than ever.

"Now, what you learned back then was that the transmission mechanism worked largely through housing. Why? Because long-lived investments are very sensitive to interest rates, short-lived investments not so much. If a company is thinking about equipping its employees with smartphones that will be antiques in three years, the interest rate isn’t going to have much bearing on its decision; and a lot of business investment is like that, if not quite that extreme. But houses last a long time and don’t become obsolete (the same is true to some extent for business structures, but in a more limited form). So Fed policy, by moving interest rates, normally exerts its effect mainly through housing."
So Krugman suggests that for the most part, corporate investment is not part of the transmission mechanism and that they key driver is housing. Here's Krugman again (emphasis mine):
"As I said then, there’s a definite change in the character of recessions after the mid-1980s. Before then, recessions were basically brought on by the Fed, which raised interest rates sharply to curb inflation, causing a slump in housing. When the Fed decided that we had suffered enough, it let rates fall again, and there was a surge from pent-up housing demand. Morning in America!

"Since then, however, inflation has been well under control, and booms have died of old age — or more precisely, they have died because of overbuilding and an excessive level of debt. The Fed is then in the position of trying to goose housing (which is the principal channel for monetary policy) even though housing may already be overbuilt (which was the point I was making, sarcastically, when I said long ago that the Fed has to create a housing bubble), and it is cutting rates from an initial level which isn’t that high. So the odds of running up against the zero lower bound are high, and recovery can be a long time in coming."
The quote above further emphasizes the housing channel, but acknowledges that after a housing bubble that has ended with excess inventory, the the Fed's policy efficacy may be reduced.

We'll revisit the housing channel in more detail soon, but first, what about the sometimes cited expectations channel? Here is blogger Winterspeak on Why the inflations expectation model is nonsense:
"Primarily, people save for things:
1. They cannot afford currently (and don't want to buy on credit)
2. Unexpected emergencies
3. Old age
4. Bequests for their kids

"2-4 cannot be moved forward and so are not inter temporal decisions the way the models treat them. Therefore, if savings are threatened, people will substitute into other stores of value such as fx or gold. They will not move consumption forward, as they cannot. This is not about consumption."
Additionally, if it turns out that there is no transmission mechanism from interest rates (the main policy lever of a central bank) to economic growth (and possible resultant inflation), then why should people alter their behavior based on the proclamation of Fed officials that they might "target" higher inflation? Target how?!?

Back to housing. We'll discuss a starting scenario (Scenario A) and four possible outcomes of monetary policy changes (Scenarios B, C, D, and E).

Scenario A (Starting Scenario) for Housing

Let's look closely at whether the housing channel EVER works (i.e., even when there has not been a recent housing bubble). Consider the diagram below of the relative quantities of different measures of housing: ownership (owned outright versus with a mortgage), vacancies (vacant versus occupied), and occupancy type (whether occupied homes are lived in by owners or renters). The diagram's format and example numbers were in part inspired by a graph featured at Calculated Risk, though I've chosen a slightly different way of showing things.

Scenario B for Housing

Now for scenario B, let's assume the central bank lowers interest rates in an attempt to stimulate the economy. The conventional wisdom is that lower interest rates motivate more people to borrow to buy houses. That's represented by the red arrow showing the size of the "Owner Has Mortgage" block expanding in the diagram below:

How do the other blocks react so that the first three columns always remain the same total height? In step 2 of this scenario, more new houses are built to keep up with this demand. But as step 3 shows us, the result is more vacant homes, since the central bank wasn't able to manufacture more people at the same time as it lowered interest rates!

What is the carrying cost of vacant homes? My cursory research suggests that in the US, property taxes range from 0.1% to 2% of a home's value (usually around 1%) and that annual maintenance costs fall in the range of 1%-4% of a home's value. Even if we very conservatively assume the average carrying cost of vacant housing were only 2% of the home's value per year, who exactly is supposed to shoulder that cost burden without putting their "excess" (not lived in or able to be rented) housing inventory on the market, thus competing with new home sales, and counteracting the effect shown in step 2?

Does this dynamic rely on highly inefficient markets, "irrrational" individuals, or sufficient time lags to induce a temporary spurt of economic growth? It seems plausible (to some degree) but not entirely convincing.

There is a situation in which it seems reasonable to assuming a growing supply of vacant homes -- during a housing bubble. The reason is that if home values are appreciating fast enough, home owners will overlook the annual carrying cost of the homes and assume that they are a profitable "investment" due to the unrealized capital gains. But housing bubbles are not the norm!

Is the inventory held tax free by home builders? Probably not -- on aggregate nationwide, they are unlikely to build excess homes if there aren't sufficient buyers in the pipeline.

Scenario C for Housing

Here's an alternate scenario to consider. In Scenario C (diagram below), steps 1 and 2 are the same as Scenario B (lower rates lead to more homes bought with mortgages), but in step 3 people are motivated by the lower mortgage rates to move out of shared multi-person housing and into newly purchased homes, reducing the average occupancy rate. So this is a scenario in which monetary policy could credibly have the intended effect on economic growth via the housing channel as more homes are built, creating more jobs and higher GDP. But how large is this dynamic? How many households have gainfully employed individuals (who would qualify for a mortgage) just waiting to move out of their friend's house or parents' basement when rates get low enough?

Scenario D for Housing

Let's consider a third scenario. After the central bank lowers interest rates, more people buy homes by taking out mortgages. But on average, owners who had no mortgage are net sellers (perhaps there is a relative portfolio shift by this demographic, on average, to other choices of assets). So in Scenario D, all that occurs is a shift in overall ownership mix between owners with mortgages and those without! There is no impact on economic growth or job creation!

Scenario E for Housing

Lastly, let's consider the possibility that lower interest rates have the effect of raising housing prices above what they otherwise would have been, since buyers obtaining mortgages would potentially qualify for a larger loan amount, allowing them to bid more for houses. To the extent that this occurs, lower interest rates wouldn't have as large an effect in lowering housing affordability (monthly payment) as the interest rate changes alone would suggest, and this scenario's diagram would look just like the original -- Scenario A... unchanged, with no meaningful effect on GDP.

In reality interest rate changes almost certainly aren't completely offset by housing price changes in this way, but there could be a partial effect along these lines!

Other Economic Impacts of Interest Rate Changes

Changes in interest rates might have impacts on other prices in the economy. For example, asset prices could change, driving changes in behavior due to portfolio and wealth effects. Savings rates might change -- for example when rates are lowered people need to save more to achieve savings goals such as for retirement, thus hurting GDP growth. Since the government is a net payer of interest (on government debt), lower rates reduce income payments to the private sector, also hurting growth. But interest rates paid between borrowers and lenders (both in the private sector) can also be impacted, causing distributional shifts. The point is that interest rate changes can have effects, but monetary policy is what MMT economists call a blunt instrument, and the overall effects are very difficult to determine.


Which scenario, B, C, D, or E is the most likely outcome of changes in interest rates with respect to the housing market? Are there other reasonable scenarios not covered in this post? Why do mainstream economists such as Paul Krugman recognize the fallacy of composition when it comes to whether lower wages will create jobs, but believe that lower interest rates have so much power over economic growth?

I would guess that in reality a mix of all four scenarios (B, C, D, & E) occurs. Given the limited macroeconomic impacts of each scenario, the bottom line is that the power of interest rates to affect economic growth and jobs appears very small. Could it really be true that the tool most readily used to manage the economy is a policy lever whose success depends as much as anything on bribing a few employed young adults to move out of their parents' basements, while at the same time exerting other contractionary effects on the economy that the mainstream fails to discuss?!?

UPDATE (same day): Moved an incongruous paragraph and assigned it to a new Scenario E, with minor rewriting. Also added one sentence within "Other Economic Impacts."

EconViz: Want to Design a Fun Educational Experience about Economics?

An excerpt from the latest post at EconViz Blog (my other economics blog):

Are you a good storyteller? Do you like the idea of creating a fun educational experience so compelling that people can’t help but learn new things about the economy and in turn share the experience with their friends?

If so, let’s talk about collaboration possibilities! Think of the technologies within EconViz’s tutorials as a sort of “palette” that can be reused in new contexts, mixed in with new content and functionality. The palette includes components such as a dynamic circular flow diagram (or any flow animation), programmable balance sheets, and potentially much more.


If this interests you, here's a link to the full post: Want to Design a Fun Educational Experience about Economics?