Don’t pay any attention to interest rates
Pay attention to the things that cause interest rates to change.
This post was inspired by a Tyler Cowen post entitled “Don’t obsess about interest rates.” I agree, and will stake out an even more extreme position.
Before doing so, I’d like to talk a bit about the identification problem in another area—wages. Suppose someone argued that rich country workers can’t compete with low wage countries like China. The counterargument would be that Germany has led the world in exports for most of the past decade, and they have some of the highest wages in the world. Bangladesh has a much bigger population, but puny exports. The counter-counterargument is that Germany’s high wages are caused by their high productivity, and that high wages are still a disadvantage in trade, holding other things constant. In the end, the debate is sterile unless we understand all of the underlying fundamentals, in which case wages are superfluous.
Here’s better analogy. Massachusetts saw a big loss in jobs in industries like shoes and textiles during the 1980s. Was the job loss due to high wages? If so, then why didn’t Massachusetts see a higher unemployment rate? Economists would use the good old comparative advantage model. The booming high tech, health care, and finance sectors created lots of high paying jobs. This raised the cost of living here, and drove out the low paying jobs. Thus our low wage jobs weren’t stolen by cheaper labor in Mexico or China, they were literally pushed out (or crowded out) by higher paying jobs in other industries. That’s creative destruction.
Now let’s consider the role of wages. Although they weren’t the root cause, it is true that high wages were a sort of transmission mechanism between the high tech boom and the loss of low wage jobs. So why do I object to people saying high wages drove out those older industries? Because wages change for many different reasons. They might change because the shoe industry became highly unionized and drove up wages, despite weakness in other areas of the economy. Or they might change because of the creative destruction process I just described—a high tech boom pushing up the overall wage rate around here. It makes a big difference which factor is behind the wage change. You haven’t really described the reason why shoe-making jobs left the state unless you can explain why wages rose.
Here’s how this relates to interest rates. Tyler Cowen links to a study by Glaeser, Gottlieb, and Gyourko (GGG), which finds that only about 20% of the housing price boom was due to low interest rates. I don’t wish to contest that study, rather I want to warn people to be careful interpreting the findings. For instance, interest rates might have fallen because the popping of the high tech bubble in 2001 led to a big drop in business investment. With lower demand for credit, interest rates (real and nominal) would naturally be expected to fall. Or the Fed might have cut interest rates. Or saving rates might have risen in Asia. If the housing industry was not directly hit by any unusual shocks, these outside factors affecting interest rates would be expected to increase housing demand, prices, and output. The GGG study is essentially saying that 80% of the rise in housing prices was due to non-interest rate factors, i.e. factors directly affecting the housing market. I suppose these could be changes in regulation, financial innovation, banks taking greater risks, GSEs, etc.
So why don’t I like hearing people talking about the portion of the housing boom that was due to low interest rates? The problem is that many listeners subconsciously connect low interest rates with “easy money.” They assume it is the Fed’s “fault.” To their credit GGG avoid this error, referring instead to “easy credit,” i.e. low real interest rates.
Economists generally accept the idea that the Fed only affects real interest rates in the short run. (Or at least I thought this was generally accepted, the last two years have caused me to reconsider everything that I assumed was generally accepted by economists.) The standard view is that the so-called liquidity effect only impacts real interest rates for a few years. In the long run, easy money will leave real interest rates unchanged, and raise nominal rates. For instance, compare the 1970s to the 2000s. The 1970s saw higher and higher nominal rates, culminating in 20% rates by 1980. The 2000s saw much lower interest rates, which ended up near 0% by 2010. The man on the street would assume that money was much tighter in the 1970s. But economists know that this is wrong, because the 1970s saw high and rising inflation, and the 2000s saw low inflation.
So here is my suggestion. Never reason from a price change. Don’t ever say “oil prices will be high next year; hence I expect oil consumption to decline.” That’s bad reasoning. And don’t ever talk about high or low interest rates causing some sort of change in the economy. Instead say something like “I believe the high tech bust helped boost the housing industry.” That sort of reasoning is consistent with the central idea of economics—scarcity. If less resources are devoted to making one type of capital good, then you’d expect more resources to be devoted to making other types of capital goods. Interest rates are merely the transmission mechanism that facilitates the “recalculation.”
Of course I am not suggesting that the tech bubble bursting caused the housing boom. GGG find only about 20% of the boom is explainable via interest rates, and the tech bubble bursting is just one of many variables that can affect real interest rates. High Asian savings rates are another factor, and Fed policy is a third. However, the liquidity effect is less significant than usually assumed, as interest rate changes far more often reflect economic conditions than exogenous changes in Fed policy. So even if low interest rates had caused the housing boom, you would still not be able to claim that a Fed easy money policy contributed to the housing boom.
PS. BTW, also avoid talk about disinflation/deflation, which could be due to either less AD or more AS. If you are worried about too little AD, then talk directly about falling NGDP. And yes, I am at fault here as well, as Bill Woolsey has noted on occasion.
PPS. I will be travelling a great deal in August, so I won’t be able to do much blogging.
Tags: housing bubble, tech bubble
3. August 2010 at 12:32
“I suppose these could be changes in regulation, financial innovation, banks taking greater risks, GSEs, etc.”
Why can’t the Fed can notice these in near real time and decide to raise rates accordingly to offset these factors?
Is you point, that they can’t negate those factors with higher rates?
If they can notice them in real time, and they can negate if they want to… they must have had a reason, and you can blame them for it – if their reason was wrong, right? So if those other factors caused “easy money” – ands Fed doesn’t negate them, isn’t it culpable?
3. August 2010 at 12:39
http://www.americanthinker.com/2010/08/paul_krugman_gives_up_1.html
3. August 2010 at 13:26
I think most who claim that “easy money” was to blame, wouldn’t focus on real 10-year treasuries as the critical variable at all. Many have highlighted the divergence between the popularity and cost of fixed vs. adjustable rate mortgages over this period. A much better indicator would be something like Sumner uses to evaluate monetary policy (expected nominal GDP growth relative to a trend or targeted to a level or something)
I find this mystifying honestly. I don’t think anyone is saying that one factor explains all of the housing boom/bust. Most sane people agree that many factors have helped explain it. The paper itself says it can account for 45% of the change in home prices as due to low real interest rates. It may not account for the majority of the price increase over the critical 2000-2005 period, but isn’t it the most significant factor the authors identified. How is that not important?
Tyler cut off the authors’ full quote but it finishes with, “However, if one is going to cherry-pick time periods, it also must be noted that falling real rates during the 2006-2008 price bust simply cannot account for the 10% decline in FHFA indexes those years.”
Of course this is a strawman. No one is or was arguing that because real interest rates fell over this period home prices should have kept rising. A proper analysis of “easy money” (like expected nominal GDP growth) showed that money was neither easy/tight over this period (until early 2008 or so). Hence, most would agree that factors other than “easy money” were the primary factors responsible for the bust.
3. August 2010 at 15:36
Scott!
Don’t leave us without your blogs.
Hole up is some motel and blog from there. Or from your car. Or from your tent.
3. August 2010 at 15:41
If you read Anthony Downs, from Brookings, or watched global capital markets, you know there was capital everywhere for anything through 2007. “Too much money chasing too few deals,” was a constant refrain.
This is caused by higher global savings rates. I am not sure of all the ramifications of global capital gluts. I sense they will recur again and again, due to high global savings rates. I suspect it means capital (supply and demand) will be cheap. If you want absolute security–US Treasuries–expect small negative real yield.
Some people feel entitled to returns, and this may come as a shock. But low tax rates in the US (hey, the top rate used to be 90 percent), and high-savers in Asia, and an aging population in Europe means capital gluts. It may also mean mad hunts for yields.
I have yet to read a smartie connecting high savings rate to monetary policy, the ramifications thereof. Sumner said he would broach the topic someday, but now he threatens vacation.
3. August 2010 at 15:49
Morgan, The Fed’s monetary policymaker should pay no attention to the housing industry–that’s for regulators. They should focus on NGDP.
Margan #2: Funny article. Thanks. I still welcome 6 inch comments. I’ll start bragging that my comments are bigger than his comments.
John Hall, I just skimmed the paper, and wasn’t trying to defend the conclusions. I was saying that even if interest rates explain some percentage of house price changes, that really doesn’t tell us anything interesting unless we know why interest rates changed. Many people assumed it was Fed policy, but other factors are much more important.
JimP. I have no laptop, and I don’t think the internet extends to Wyoming.
3. August 2010 at 16:55
if you (or another commenter) could help me understand i’d be greatly appreciative. after following this blog i’m leaning toward agreement with the idea presented in this post but i’m not sure if i fully understand it.
i’m a bit confused how this view of interest rates squares with the Fed exogenously changing them. i understand the natural rate of interest and how it may differ from the Fed’s (at least i think i do), but if the Fed can decide to change rates almost as they please and assuming educated businesses/people really do react, doesn’t that mean their decisions/target interest rates can’t be ignored and have to be obssessed over? basically it seems that since the Fed will decide a target interest rate the only answer to “why” is “because the Fed said so”. or are you implicitly arguing that even if the Fed targets an interest rate the market will stick with the natural rate (or whatever rate) regardless?
thanks a lot for any help, blogs like this are a goldmine of info!
3. August 2010 at 17:54
Scott, good lord, Greenspan just said “if housing prices fall it will be a catastrophe!” All they have been thinking about is the housing industry.
Why do you insist this is about other stuff? Housing is the only deflater, it is the crisis – it the elephant in the room on every post you write. It’s like talking about politics and not mentioning Democrats.
Thought exercise, (not something I’m in favor of):
We have 20M Public Employees. Let’s say they are all being paid 2x as much as they should be in a purely competitive world… on the premise that government jobs are really make work jobs and therefore they should pay less than market rates including pensions and benefits.
So, what happens if we start to void their contracts as they expire, and decide from now on we’re going to buy twice as many and pay them all half as much?
Suddenly the unemployed are all happy to vote for the party (Repubs)that promises these changes… maybe the argument is that the structurally unemployed (from these technology gains) all deserve to feed in the trough of our largess (taxes) along with all the other public “servants.”
What are the effects?
1. Unemployment hits 5% or less, right?
2. I assume rents go down, because rents track to pay, not mortgage value.
3. But GDP stays basically the same?
4. Wages go up in private sector?
The reason I’m against it is I don’t think we need 40M people voting for higher taxes on the rich. But apart for that, what bad about it?
3. August 2010 at 18:36
Benjamin, Yes, high saving rates may be a factor here.
Does this mean equities are a good long term investment? (The logic is that high Asian saving rates might keep T-bond yields low for years to come.) I hope so.
ltamaye, That’s a difficult issue to explain, but the Fed doesn’t really have the sort of control over rates than many people assume it has. From day to day it can change rates, but over long periods it must adjust rates to reflect the condition of the economy, or else the price level becomes indeterminate. For instance, of the Fed pegs the interest rate at a fixed rate, the price level will fly off to zero or infinity. We don’t see this, because the Fed cuts rates when the economy slows, and vice versa. So the Fed is actually mostly taking its marching orders from the market, but it doesn’t look that way.
Morgan, I didn’t say the Fed doesn’t look at housing, I said they shouldn’t look at housing.
Yes, let’s reform the public sector, but it’s not as easy as you make it seem.
3. August 2010 at 19:42
“Upside down…round and round”
http://capitalgainsandgames.com/blog/stan-collender/1885/bond-vigilantes-are-now-deficit-cheerleaders?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed:+CapitalGainsAndGames+%28Capital+Gains+and+Games+-+Wall+Street,+Washington,+and+Everything+in+Between%29
3. August 2010 at 19:45
perfect, thanks Professor. i just wanted to be sure i was reading your posts with the right kind of framing.
3. August 2010 at 20:22
Could the housing bubble be likened to when the “price level [flies] off to … infinity”? Suppose that process were more localized in terms of certain goods, and more drawn out in time, than a single variable (the price level) would imply.
That was essentially how I imagined the fed’s role in the housing boom.
4. August 2010 at 08:02
“So here is my suggestion. Never reason from a price change. Don’t ever say “oil prices will be high next year; hence I expect oil consumption to decline.” ”
Scott, can you elaborate on this? I don’t understand how you would like the reader to square this with the insights you’d like us to draw from efficient markets.
4. August 2010 at 09:47
Scott – another great one. I seem to enjoy you most when hitting on this theme.
I also share Morgan’s first question. When government ran fixed exchange rates, it didn’t matter what the outside pressure was forcing them up or down from that level – recession, inflation, speculation, demand for products – the monetary authority fought to maintain the targeted exchange rate. It was a policy failure if they didn’t respond. This is just a price, but it was the policy lever. Letting outside forces control your policy lever is bad business. Why, when the Fed is nominally (forgive the pun) targeting interest rates, is it not a policy failure when it doesn’t take steps to counteract outside forces? You can argue there’s a deeper policy failure in targeting exchange and interest rates at all, but that doesn’t negate there being a policy failure in not responding.
You posted earlier that silence is a policy stance. If the Fed is silent when outside forces lower interest rates, is it not responsible for the further effects of the interest rate changes? It could be that the proper response is looser monetary policy (recession lowered the interest rate) or tighter, but doing nothing is still doing something.
4. August 2010 at 13:33
They persist…
http://www.washingtonpost.com/wp-dyn/content/article/2010/08/03/AR2010080302685.html
4. August 2010 at 18:46
Wow. I would not be surprised to find almost perfect correlation between Federal Reserve interest rate policy and mortgage rates during the whole decade. It was only talked about on the likes of CNBC incessantly.
Housing affordability was sold by the industry as the monthly payment of those teaser rates “because housing prices never go down”. Greenspan also advocated subprime mortgages and declared there was no housing bubble. There’s no way Federal Reserve interest rate policy is only a 0.2 weighted variable.
Did you even notice that the Federal Reserve stopped publishing M3 money supply data early in the decade? And don’t forget the HELOCs used to flip additional houses.
http://thehousingbubbleblog.com/index.html
was years ahead of the curve. Years after the event and you are still way behind the curve. Sure there are other heavily weighted variables like GSEs, no doc liar ninja loans broker originations fraud, securitization of mortgages and derivative insurance schemes, but none of those variables would have come into play without the Federal Reserve setting interest rates below what a free market would have set them at.
6. August 2010 at 05:54
Marcus, Yes, so much for “bond vigilantes.”
ltamaye. You’re welcome.
MarkWash. No, the price level refers to the average price of all goods, not just one good.
OneEyedMan, Oil prices can fall because of low demand or high supply. The price alone tells us nothing. That is consistent with the EMH. Here is where the EMH comes in. If forward oil prices fall, then this suggests that future oil prices are likely to also be lower.
Your question raises a good point. Deflation could occur due to more AS or less AD. I should really stick to NGDP, as it is a better indicator of low AD than deflation. So I often make the same mistake.
D. Watson, Thanks, I agree that Fed passivity is just a much a policy as anything else. Whatever happens to nominal aggregates reflects policy in some way.
Thanks Marcus.
Monximus, I think you missed the point of the post. I wasn’t trying to argue that 20% of the housing boom was due to interest rates, I was arguing that interest rates mostly reflect market forces, not Fed policy. And low interest rates often reflect tight money.
10. August 2010 at 11:19
Hmm, announcement came in. Looks like “Bad” it is then. I guess it could have been worse?
12. August 2010 at 23:31
[…] interest rates are close to 0%, but you should ignore interest rates. Lots of other things impact monetary policy apart from the headline interest rate, […]
23. August 2010 at 11:51
Pragmaticon, yes, it could have been worse, but it was still pretty bad.
23. January 2011 at 06:26
[…] by Scott Sumner　 // Don’t pay any attention to interest rates（3. August […]
28. March 2011 at 03:22
[…] 2. Krugman’s third sentence (“It’s a process that works like this: lower prices -> lower demand for money -> lower interest rates -> higher spending.”) need not be the dominant causal mechanism when so many variables are changing. Scott in particular might think that interest rates are not so important. […]