A wee bit more on spending, borrowing, and earthquakes

Regarding Brad’s response post. I view our disagreement as such.  In my view, a low rate of interest on Treasury securities indicates a high chance that the investors will be paid back.  That is all it means.  It is not a market signal that the government output should be produced.  Investment is a normal good and post-earthquake the world is poorer, and riskier, so optimal investment can either go up or down.  The interest rate on debt is not a sufficient statistic summing up all of the conflicting “investment should go up” vs. “investment should go down” forces in our thought experiment.  It is not summing up real consumption risk or output valuations, only whether the government will pay back investors (None of this, by the way, need deny that there might be other, non-earthquake reasons for expanding government spending).

Here is an exaggerated but clarifying example.  Under Ceausescu, Romania could borrow at low rates, but that just meant the government would pay you back.  It didn’t show the outputs were properly matched to risk or to the wealth of the society.

Also note that private firms — whose output has to face a real market test — invest in a manner which is notoriously insensitive to the real interest rate.

This debate needs to revisit Scott Sumner on the dangers of reasoning from a price change.  Don’t consider the price change without also working through the reasons for the price change.  Brad is citing the price change alone.

Another good rule of thumb: when you have a complex problem, and one economist (me) claims “the right answer is indeterminate,” that economist (me) is usually right.

Addendum: Greg Ip has relevant comments, as does Ryan Avent.  By the way, a related response is to take seriously Brad’s endorsement of the liquidity trap idea.  Then a negative supply shock should be expansionary.  What is the implied prediction for interest rates?  Did it come to pass?  There is now also a possible “positive wealth effect” argument for expanding government investment but does Brad wish to make that?  Isn’t the market price response to the earthquake evidence that we are not in a liquidity trap?  I stress, however, that this is a separate line of response.

And a response from Brad.

Comments

Certainly one test all respondents should apply is to ask themselves whether they would support the antisymmetric argument if U.S. Treasury yields were to increase.

My estimate is that Mr. De Long would have found reasons not to espouse the opposite view in the opposite scenario.

Sammler's comment is apt. Municipal yields (and spreads) on tax-free General Obligation bonds (for largely technical reasons) have widened considerably in the last six months. Does DeLong advocate for continuing fiscal contraction in states and municipalities? How does he really want his "fifty little Hoovers", if one reasons purely from price changes?

His response, I imagine, is that precisely because seignorage allows the US to enjoy an exalted position of reserve currency issuer, we should utilize that to minimize aggregate demand & employment dislocations, till the market, through USG bond yields, signal otherwise. Hence munis do not make the cut, & so wouldn't, I hazard, any other sovereign countries unless, like Japan, there was a large pool of risk-averse domestic savers who had an insatiable appetite for their local currency-denominated debt.

There are certain categories of government expenditure where the cost of funds should be the key consideration for whether or not to proceed, such as rebuilding bridges. Assuming that a given project will need to be completed at some point, it should be finished when the cost of completing the project is low. Any project whose timing can be controlled and whose cost/benefit analysis is sensitive to hurdle rates (aka has a long payback time) should be executed today. This includes most infrastructure project and fundamental research grants. The reason for the price change has no impact on this particular line of reasoning.

The view that the risk premium has increased encourages many to switch to Treasuries. This lowers interest rates. The government could take steps to lower the risk premium through pro-growth policies that encourage private expansion. The lower cost of servicing government debt could be passed on to the citizens in the form of lower taxes.

But Brad De Long sees every event as an opportunity to expand government.

In the case of Japan, they will need to make substantial investments in their power grid. Is it an advantage to have the government build the new infrastructure? By his comments, Mr De Long seems to think that the government is the only option.

The market's response doesn't mean that we're not in a liquidity trap: it's a result of an increase in the price of risk, which exacerbates the liquidity trap even as the supply shock ameliorates it. It should also be noted that, from the US point of view, the "supply shock" is not, for the most part, really a supply shock but a positive demand shock. The damage to US capacity from an earthquake in Japan is minimal. What the US will see is the effect of an increase in Japan's net import demand (which is to say, Japan's trade surplus will fall). From the US point of view, the net income effect (down due to the price of risk but up due to increased net foreign demand) is ambiguous and likely quite small, but we are observing a decline in the price of current government spending. Technically, the right answer is still indeterminate, but the preponderance of the evidence falls on Brad's side.

"Also note that private firms — whose output has to face a real market test — invest in a manner which is notoriously insensitive to the real interest rate."

A "real market test" can just as easily - and I would think almost always is - manipulated by state intervention or other monopolized processes. This can often shield firms from the insolvency they may otherwise face. Notably, the reduction of mark-to-market regulations removed the influence of government regulation - while also distancing investment banks from real market valuations of their assets.

It might be wise to consider that there is always incentive to shield a firm from the influence of uncertain conditions which might bring them "just desserts." Probably just important to remember is how much more effectively a private market process will be at achieving this state of "invulnerability" - they are more efficient at accruing profit, why not also at influencing the state and relevant firms (including ratings firms) to their own ends?

Low interest rates for US Treasuries indicates US taxpayers and Japanese and German taxpayers and oil rich dictators are willing to pay for bigger US government, because paying a low price for US government services and product, and then buying US government debt is preferred to paying taxes for the full price of government services, because with debt you have the illusion you can profit from government services you aren't paying for but are liable for. The Saudis would rather buy US government debt instead of having oil taxed enough to pay for the US government services provided to the Saudis to protect their dictatorship from democracy.

But this is like the Japanese buying government debt instead of paying higher taxes to fund the government they want.

And this shift in the way big and bigger government is paid for is the fruit of the Reagan revolution who proved that switching to debt from taxes to fund bigger government makes it easier to sell your big spending to the public. How many Americans would have agreed to higher taxes to pay for Reagan's massive military spending? FDR, Eisenhower, JFK, and LBJ imposed enough in taxes to pay for the military, but they had to sell their military spending to the public. FDR's CCC and WPA created jobs while making young men healthy for the military. Eisenhower built highways for defense that allowed for people to use their cars. JFK created the high tech industry and NASA as a spin off from his military spending. LBJ hiked taxes to pay for Vietnam. Ooops, hiking taxes to pay for war was taken off the table for at least half a century.

Let's destroy Washington DC and see what happens.

It appears that Brad DeLong thinks that if price signals indicate that we should pile bananas on our roofs until they collapse, that is what we “should” do. The last time I checked, “should” isn’t part of economics. “Should” is part of some other discipline… in this case, hackery.

Quote:
> post-earthquake the world is [...] riskier

earthquakes are not i.i.d. but strongly serially correlated, therefore Japan (if not the world) is less risky after an earthquake.

Side remark: I fully agree with Andy Harless's comment above.

. Investment is a normal good and post-earthquake the world is poorer, and riskier, so optimal investment can either go up or down. The interest rate on debt is not a sufficient statistic summing up all of the conflicting “investment should go up” vs. “investment should go down” forces in our thought experiment. It is not summing up real consumption risk or output

It should also be noted that, from the US point of view, the “supply shock” is not, for the most part, really a supply shock but a positive demand shock. The damage to US capacity from an earthquake in Japan is minimal.

Out of curiousity, how did you arrive at this result? While obviously US capacity stock has not suffered from the earthquake in Japan, to the extent that production processes in the US depend on supplies from affected areas of Japan, then US capacity might well be affected. Furthermore, while demand for US goods that assist in rebuilding Japan presumably will go up, meanwhile demand for US goods from whatever the Japanese are now not spending money on will presumably fall.
It may be as you say, that the earthquake in Japan is a positive supply shock from the US's point of view, I just would like to see the evidence you used to get to this position, as I may find it useful elsewhere.

We're confusing costs of a project with financing of a project. Yes, lower interest rates indicate that the financing costs of borrowing is lower, so there should be a marginal shift of not increasing taxes in favor of debt financing. Lower interest rates say very little about what the optimal level of government services should be. However, as Sohier notes above, one could argue that certain infrastructure projects might be accelerated because the financing costs of debt to support the project are lower--but this is just a timing issue--and are easily handled at the local and state level.

Isn't what's missing here an acknowledgement of, "All else held equal"? All else held equal, yes, cheaper borrowing (cheaper anything!) means we ought to consume more of it.

But is all else held equal here? If I determine I'm not saving enough for retirement, but my credit card interest rate drops, should I spend more now? Not necessarily.

Comments for this post are closed