Don't borrowing-constrained households spend more of a tax cut??

One of the standard arguments against the Ricardian Equivalence Proposition is that some households are borrowing-constrained. They want to borrow and spend against their future income, but can't find anyone to lend them money. So when the government gives them a tax cut, financed by higher future taxes, the government is effectively lending them the money they had previously wanted to borrow and spend. So borrowing-constrained households will spend most or all of a tax cut.

Now I think I keep hearing what seems to be the opposite argument. Mark Thoma, for example:

"This brings up a point about tax cuts I've been meaning to make (again). The effectiveness of tax cuts depends, in part, on how hard the recession hits household balance sheets. In a recession where balance sheets are relatively unaffected, a tax cut may very well translate into spending, and do so fairly quickly.

But when balance sheets are hit hard, the result is different. In this case tax cuts may be used largely to rebuild balance sheets - to recover what was lost - rather than for new spending. Thus, in this recession the stimulative effects of tax cuts may not have as large of an immediate effect as in the past..."

Now, it's not obvious that when Mark talks about balance sheets being hit hard it means exactly the same as households being borrowing-constrained, but they should be related. If I have a bad balance sheet, with high debts relative to my income and assets, it should be harder for me to borrow to finance consumption. So if household balance sheets have been hit hard, and some of them have been hit hard by the fall in asset prices, more households should be borrowing-constrained.

And I can see no obvious mistake in Mark's argument either. If you have more debt, relative to your income and assets, than you feel comfortable with, it seems entirely plausible that you would use most of any increase in income to pay down your debts, rather than spend it on increased consumption.

So how do we reconcile these two apparently plausible arguments, that nevertheless point in opposite directions? If a household's balance sheet deteriorates, and that household becomes more borrowing constrained, is that household more or less likely to spend a tax cut on increased consumption?

I am undecided between two resolutions to this apparent contradiction, so I am going to present them both:

1. Mark is wrong, and the standard argument is right. Households whose balance sheets have been damaged by the fall in asset prices will be borrowing-constrained, and will be more likely to spend a tax cut. It is true that a household with a bad balance sheet will have a lower average propensity to consume, because they can't borrow, and need to make the mortgage payments. But their marginal propensity to consume will be higher, because a $100 tax cut enables them to increase consumption by $100, relative to before the tax cut, while still meeting the mortgage payments.

2. Mark is right, and the standard argument is wrong. A borrowing constraint is not an absolute constraint, that is either binding or not binding. It binds progressively, and the more it binds the more precarious is your way of life. A bad balance sheet is more like an additional demand on your disposable income, that competes with consumption for your marginal dollar. So a $100 tax cut may be divided between those two demands, and a large part of it may be used to pay down your debts more quickly.

I'm actually leaning towards the second resolution. But I really wish I could get my head clear on the implicit underlying assumptions that could make it valid.

Which means it's time for me to go canoeing! Back in a few days.

Comments

[anon]

Apart from the effect discussed, if tax cuts are funded with government borrowing (dissaving), then some non-government sector MUST "improve" its balance sheet by buying government bonds (saving).

[Nick Rowe]

anon: that's true in the new equilibrium. But starting in the original equilibrium, if there are two sectors, government plus households (ignore firms and foreigners), then if government tries to borrow, but households want to spend the tax cut, then income expands, until households eventually save enough extra to match the amount the government wants to borrow.

[AndyfromTucson]

I think it has to do more with expectations for future income than it has to do with current balance sheets. A borrowing-constrained household is a household that is pretty confident about its future income, otherwise it wouldn't be trying to spend it. If something happens to change those expectations about future income for the worse (say a dramatic recession), then the household is perceive they have less future income to spend today, and they will cut back on current spending. In other words, gloomy expectations about future income will decrease the marginal propensity to consume independent of the household's current balance sheet. Or put another way, I think households have a mental balance sheet that not only includes current assets and liabilities, but also expected future income and expenses, and when the prospects for future income dim it is perceived as a current loss of wealth, which results in a current reduction in the propensity to consume.

[bill woolsey]

When households "improve their balance sheets," what are they doing?

Paying down debts? Then what do those receiving the debt repayments do with the money?

Accumulating financial assets? What do those selling the assets do with the money?

Because net debt is zero, we cannot all improve out balance sheets at once.

Of course, monetary disequilibrium--an effort to accumulate more money with a given quantity--
can result in less spending on everything else--including especially currently produced output.

So, one might say that falling asset prices reduced net worth, and so people want to accumulate more assets and pay down debt. So, why is this a problem?

It is a problem if they all want to purchase short term, low risk assets (like T-bills) and have bid their yields down to zero or their abouts, and so any remaining shortage of these seucrities gets shunted over to money. And then, nominal income must fall to bring the demand for money down to the quantity.

Now, explain to me how "waiting" for people to fix their balance sheets is really going to help?

Nominal income falls. While those that continue to be employed may well improve their balance sheets, those earning less aren't doing so.

It is like saying that there is a shortage of money, so people spend less and accumulate more money. After a time, they will have accumulated more, and then the econmy will recover... Right.

My comment is not about the effectiveness of fical policy. I think it is pretty clear that it will work by increasing the supply of short term, low risk assets that are in shortage at yields near zero, and so reduce the demand shifted to money, and so reduce the monetary disequilibrium.

It will actually work better than open market operations where money is created through the purchase of the exact same low risk, short term assets with a shortage at the near zero nominal yields.

But this notion that the economy must weight to repair balance sheets.. just seems wrong.

Those househods that are in debt can save and pay down debts. Those households that are net creditors can use the funds received to purchase equities. The result repairs everyone's balance sheet.

Firms that are overleveraged can pay down debt. Firms that receive the repayments can purchase stock or buy capital goods.

My prefered solution to the actual problem (which isn't balance sheets in need of repair, but a shortage of money) is a negative nominal yield on short term, low risk assets, so that they clear.

But, if that can't be managed, then the solution is quantitative easing, which almost certainly requires that the central bank purchase longer term, higher risk assets.

But, fiscal policy should work too.

[Beezer]

I'm with Thoma on this. With the economy undermining confidence, households reduce their borrowing. They reduce debt first.

With everyone knowing someone who's laid off, folks simply stop spending, unless they absolutely have to spend. The longer this recession continues, the more ingrained will be this propensity to save and to reduce consumption.

[pointbite]

If you're trying to stabilize demand with government intervention, his argument sounds like "raise taxes in a recession" and "cut taxes when the economy expands". Did I miss something, or is that just a terrible idea?

[Patrick]

pointbite: Yeah, you did miss something. That is not at all what Thoma is saying.

[Matt Nolan]

AndyfromTuscon,

What you said.

I agree Thoma's argument is based on the fact that we have had a negative wealth shock, so debt levels are higher than household's want them to be. In essence, there is no credit constraint at this point (as banks aren't going to stop people paying down debt) so any (appropriately sized) transfer between future and current income (which a tax cut is) will be subject to complete ricardian equivalence and won't have any stimulatory impact.

[Account Deleted]

Dear Nick,

You raise an interesting question.

1 Is there a distinction between "borrowing constraint" and "balance sheet stress"?

If I'm trying to rebuild my balance sheet, it means my desired debt level is lower than my actual debt level. For example, I might want to reduce my debt/asset ratio from 80% to 40%. Any cash I get will be directed towards that goal.

If I'm under a borrowing constraint, I would like to increase my debt level, but cannot - perhaps because I have no collateral acceptable to a lender, or I don't have a steady income. In this case, my desired debt level is higher than my actual debt level. I may be happy to increase my debt/asset ratio from 40% to 80% - perhaps to get a new car - but I cannot get anyone willing to lend me the money. Any cash I get would be directed to the goal of buying the new car.

2 There is a time for everything, a time to repay debt, and a time to increase debt ...

In "normal times" when unemployment is low, most households and firms have a balance sheet at a debt level that they are comfortable with. However, some households (and possibly firms) may be under a borrowing constraint - perhaps because their income is volatile or because they have no collateral acceptable to lenders. Therefore, perhaps under normal times, the "standard argument" applies.

If for some reason the economy falls into a "balance sheet recession", many households and firms have a balance sheet at a debt level that they are not comfortable with. In such times, Mark Thoma's argument would prevail over the "standard argument".

Does this reconcile the "standard argument" with Mark Thoma's argument?

Cheers,

Kien

[Neil]

Mark's argument presupposes that the average household knows or cares what their balance sheet looks like. Most people's accounting skills roughly give them enough knowledge to figure out if they make enough to cover the monthly payments on a purchase...and a lot of people fail even at that. Knowing that they owe more on their house than they can sell it for is not likely to affect their purchasing decisions.

[Lord]

As Kien notes, one is constrained by being unable to borrow as much as they want, the other is constrained by already having borrowed more than they want. They don't actually have to know their balance sheets, only whether they find it easy or difficult to pay the debt. That is far less than the perfect future knowledge Ricardian Equivalence assumes.

The problem with Ricardian Equivalence is it assumes initial government action is exogenous but all future action in response are endogenous. Somehow people can anticipate all future fiscal policy from passage of a bill, but are utterly oblivious to the possibility of its passage in the first place. This is irrational unless it was totally unexpected. Certainly by now, people should expect deficits during recessions even if they expect surpluses during growth (though they might soon come to doubt that from experience). Can even something such as a war be totally unexpected?

[Nick Rowe]

Kien: I think you are onto something. I'm still getting my head clear on this.

But there are three questions:

1. What is the effect of bad balance sheets on consumption?

2. What is the effect of a (temporary) tax cut on consumption?

3. What is the effect of bad balance sheets on the effect of a tax cut on consumption?

It's that third question that's at issue.

C= C(T,B). It's not dC/dB, nor bC/dT, but the sign of the cross-partial d^2C/dBdT.