Monthly Archives: November 2014

Strong Assertions: Reply To A Comment [Update 2]

I don’t generally publish comments and reply offline via email but this one needed one on the blog.

Winterspeak commented on my previous post Strong Assertions:

Deliberate obtuseness ramanan?

A 15% interest rate will certainly reduce borrowing as a first order effect, but it will also have another first order effect which will move AD in another direction. You know what this is, why not address it directly? And then why not respond to Mosler’s primary point directly as well?

I don’t think warren’s problem is that his writing is too simple.

My response below:

Deliberate obtuseness? My post was clear that the assertion that economists have it backward is quite wrong and misleading. I do in fact mention the effect of higher interest rates because of interest payment on government debt. In my post, I said

Finally the point about interest income on government bonds: it is true that if interest rates are higher, the private sector is receiving more income from the government and this is one factor to consider among all factors which affect aggregate demand. But there is no reason to assume that this effect is always higher.

In stock-flow consistent models, one sees the long run output depend positively on interest rates. But short term, this effect isn’t always positive except in simple pedagogic models.

My example of 15% was not really purely academic. Such an experiment happened in the UK in the 70s where interest rates were raised sharply and it led to a contraction of output. Several firms had to close down because of a rise in debt burden. The full story is in Nicholas Kaldor’s book The Scourge of Monetarism. I suppose Winterspeak thinks Monetarism cannot be a scourge. There were additional effects as well. The exchange rate appreciated  and led to a rise in imports contracting domestic demand even more than via other factors.

One cannot simply say that a rise in interest rates will lead to a rise in interest payments on government debt and that hence domestic demand and output will rise because of this. Suppose the government debt is 60% and let us say the average interest rate on government debt rises by 2% initially. This will lead to an additional interest payment of 1.2% to the private sector. This does not increase output by 1.2% automatically. It depends on the interest receivers’ propensity to consume. If this is say 0.2, the first order effect is a rise in output by 0.24% only. (Higher order effects are via income/expenditure multiplier process). However, borrowing also depends on the interest rate. Suppose fixed capital formation by firms and households reduces by more than 0.24%, the latter has had a bigger negative effect than the positive effect of the former.

So the effect depends on interest elasticity for borrowing and propensity to consume from interest income. But that is not all. A rise in interest rates may also lead to a fall in asset prices and which has wealth effects on economic activity. There are other complications as well which I do not need to go into in detail because my point is made. In many countries, a lot of households took various kinds of mortgages which have amortization schedule highly sensitive on interest rates. If interest rates are raised, their monthly payments will increase leading to a lower consumption. Of course, it can be argued that the interest paid is income to some other economic unit, but one needs to look into who the interest receiver is, how their behaviour and so on.

Update

After I posted this, I received a comment again from Winterspeak:

Please.

A 15% FFR will impact more than income from the Government. And the only people assuming that a “rise in interest rates will only have one effect” are you and Monetarists. Talk about being overly simplistic.

I recommend leaving strawmen out of it and focusing on the meat of the argument.

Puhleeze!

Winterspeak tells me of being overly simplistic and attacking a strawman.

But look who is overly simplistic here. Winterspeak simply announces that a 15% rise in interest rates will have more impact than the income from the government. Clearly he has not understood much. My post talked of the propensity to consume of the interest earners and also the mutliplier effects of this. There is no reason that the effect of this (including multiplier effects) is greater than 1.

Plus Winterspeak seems to completely  ignore the negative effects on borrowing: proving my point. Ignoring intermediate consumption, the gross domestic product is the same as output which is (in a simple closed economy model):

C + I + G

where C is household consumption, I is private fixed capital formation and G is “pure” government expenditure (which doesn’t include interest payments on government debt) i.e., government consumption and fixed capital formation.

While C may rise because of higher interest earned by households because of higher interest income. can fall more because of high interest rates. It is also not clear if C will necessarily rise. This is because if households have large liabilities (such as mortgages), their disposable income will fall due to a rise in interest rates (and hence interest payments) and hence consumption as well.

Update 2

After I wrote the above, I received a patronizing comment by Winterspeak:

Great — you’re slowly getting closer.

So what did Warren actually say and why? Or, in other words, what is the logical next step from your (third) post?

Let me argue again. Let us use subscripts 1 and 2 for time periods.

Initially the GDP is

C1 + I1 + G1

Now interest rates are raised to 15%. The GDP is

C2 + I2 + G2

With pure government expenditure remaining the same,

G2 = G1

Interest expenditure of the government is not counted in production. G stands for government consumption expenditure and expenditure on fixed capital formation, not total government expenditure. (Ignoring changes in inventories for simplicity, both for firms and the government). Of course, the interest income should appear somewhere, and it will make an appearance in the consumption function.

Fixed capital formation is assumed to depend on interest rates, so

I2 < I1

Consumption depends on income, holding gains and previously accumulated wealth and propensities to consume. Propensities can depend on the type of income (compensation of employees, interest income, income via dividends) and so on.

Less fixed capital formation implies firms hire less. This means compensation paid to employees is lesser than before and also since fixed capital formation of firms is also an income flow for firms as a whole, a reduction implies less profits and dividends paid to households.

Hence, despite households receiving higher interest income on government debt, their total income is likely less than before with interest rate at 15% and hence,

C2 < C1

This implies:

C2 + I2 + G2 < C1 + I1 + G1

Which was I intended to show.

Funnily, Winterspeak puts me in the same position of Monetarists but it is him who is being a Neo-Fisherite here. (Referring to Krugman’s terminology is not an endorsement to his monetary economics.)

Strong Assertions

In a recent article (from last month), Warren Mosler makes strong claims. He says:

I reject the belief that economy is strong and operating anywhere near full employment. I also reject the belief that a zero-rate policy is inflationary, supports aggregate demand, or weakens the currency, or that higher rates slow the economy and reduce inflation.

What I am asserting is that the Fed and the mainstream have it backwards with regard to how interest rates interact with the economy. They have it backwards with regard to both the current health of the economy and inflation, and, therefore, their discussion of appropriate monetary policy is entirely confused and inapplicable.

Furthermore, while I recognize that raising rates supports both aggregate demand and inflation, I am categorically against raising rates for that purpose.

The problem with the mainstream credit channel is that it relies on the assumption that lower rates encourage borrowing to spend. At a micro level this seems plausible- people will borrow more to buy houses and cars, and business will borrow more to invest. But it breaks down at the macro level. For every dollar borrowed there is a dollar saved, so any reduction in interest costs for borrowers corresponds to an identical reduction for savers. The only way a rate cut would result in increased borrowing to spend would be if the propensity to spend of borrowers exceeded that of savers. The economy, however, is a large net saver, as government is an equally large net payer of interest on its outstanding debt. Therefore, rate cuts directly reduce government spending and the economy’s private sector’s net interest income.

I don’t understand why as a heteredox author, Mosler simplifies his analysis so much.

This post is not about discussion about whether it is time to raise interest rates in the United States. It’s not the time. It is about Mosler’s claim that monetary policy works opposite to what is usually assumed.

This oversimplification can easily be debunked. If the central bank raises the short term interest rate to say 15% from 0.25%, it will obviously lead to a reduction in borrowing. Firms will reduce investment and stock building as higher rates will require them to pay higher interest and the expectation that the economy will slow down will dampen production. Households may postpone plans for purchases of new houses and take out lesser loans and those with existing loans on floating interest rates are likely to reduce consumption when faced with a higher debt burden because of higher interest payments. Further, raising rates from say 0.25% to 15% may bankrupt a lot of firms because interest payments on debt may become very high. There are also feedback effects: a slowdown of output will lead to higher unemployment and less consumption and so on. It can be argued that interest payment by one economic unit is income for another so one needs a model to see how all this works: complications such as consumption propensities of interest payers and receivers.

Of course this effect may not be so strong if the short term interest rate is raised from 0.25% to say 2% but asserting that there is no effect and that the effect is exactly the opposite is too simplified a claim.

Does that mean that rising short term interest rates is always accompanied by a lower output? No. Monetary policy is only one channel. It is possible that while the central bank is raising interest rates, other things that affect aggregate demand conditions are working to raise output. For example, the government may be raising pure government expenditure while the central bank is raising rates, or exports are rising.

Now reconsider Mosler’s point quoted above:

For every dollar borrowed there is a dollar saved, so any reduction in interest costs for borrowers corresponds to an identical reduction for savers. The only way a rate cut would result in increased borrowing to spend would be if the propensity to spend of borrowers exceeded that of savers.

Not sure what that means. Dollar is not fixed in quantity. Further an economic unit may be both a net borrower and a saver. To see this think of a simple example: Your disposable income is $1mn, your consumption is $200,000 and you borrow $4.2mn to buy a house for $5mn. Your saving is $800,000 and your net borrowing is $4.2mn. You are both a saver and a borrower. Also, I am not sure how “propensity to spend of borrowers” means, as long as one is talking of borrowing to not make purchases of financial assets: all that is borrowed is spent, so this propensity is always equal to 1.

Perhaps Mosler has in mind the debt/income ratio. In the above example, your debt/income has risen but it isn’t necessarily the case with firms as investment is self-financing. Firms may borrow more in response to a fall in interest rates. But this needn’t cause a rise in debt/income as investment is also a source of income for firms as a whole. So firms’ debt/income may actually improve.

Mosler discusses the net lending of the private sector when he is talking of “net saving” (saving less investment expenditure), which is identically the net lending. Even if the private sector as a whole sees a deterioration of their net lending position it isn’t necessarily problematic in the short run. There is no reason that this is a constant relative to output or income which Mosler implicitly is assuming.

Finally the point about interest income on government bonds: it is true that if interest rates are higher, the private sector is receiving more income from the government and this is one factor to consider among all factors which affect aggregate demand. But there is no reason to assume that this effect is always higher.

In stock-flow consistent models, one sees the long run output depend positively on interest rates. But short term, this effect isn’t always positive except in simple pedagogic models.

Mosler gives the example of Japan to show what he says vindicates him. Low rates in Japan hasn’t helped Japan. The analysis is oversimplified because output depends on so many other things than monetary policy. There is no need to make simplistic assertions. Heteredox economists will be be seen as simpletons because of analysis such as that of Mosler. Mosler’s idea of writing was perhaps to stress the importance of fiscal policy and that mainstream economists underplay the positive role of fiscal policy and exaggerate the role of monetary policy. It can be said directly instead of claiming “the mainstream have it backwards with regard to how interest rates interact with the economy.”