Saturday, February 21, 2009

I. O. Me

Twenty-five years ago, I wrote an exam essay explaining why public debt is fundamentally different from private debt. It’s time to dust off that lesson again – not to say public debt is good, but to weigh in against glib analogies to household overspending.

The biggest difference applies to the 72% of federal debt held by Americans – all those Treasury bills and bonds in U.S. investment portfolios, safe deposit boxes and grandkids’ sock drawers. We hold both the I.O.U.s and the things we bought with them. With private debt, the borrower buys stuff, and someone else holds the I.O.U. The borrower has stuff plus debt. The lender has less stuff plus an I.O.U. With domestically held Federal debt, we have the stuff, plus debt and the I.O.U.

This is what economists mean when they say public debt is money we owe ourselves. Our debt is offset by our securities – plus whatever we bought with the debt. Even intergenerationally, this difference still holds. Future generations will have debt plus offsetting securities plus the highways and technological innovations and well-schooled kids. It’s so different from private debt that it’s bizarre and hard to wrap our brains around (which might be why so many people don’t).

The second biggest difference is that the federal government does not have a capital budget. That is, our annual deficits include the full cost of investments in the future for future benefits, such as building levees and hospitals or fighting wars to protect our values. That’s like individuals buying cars and houses cash-in-full.

The reality is that we usually make our budgets by comparing our annual income to just that year’s monthly payments for our house and car and other expenses. A person with $30k left over after all other expenses isn’t limited to buying a $30k house, but rather a house whose mortgage payments add up to $30k a year. After all, why should we pay for a house all in one year when we'll be using it a little bit at a time over 20 or 30 years?

When the federal government buys a house, the full cost hits a single year’s budget. Running a deficit and covering the deficit with bonds is how the government takes out loans and mortgages to pay for long term investments and for durable goods. Of course, there’s a temptation to run deficits to pay for others things, too, that may not have long term payoffs. But that doesn’t change the fact that investments hit current budgets and current deficits.

Finally, the 28% of debt that is foreign owned does not give the foreigners any control over the US. Treasury bonds are not stocks, which confer ownership. Nor are US debts collectable in the way that private debts can be collected by seizing assets and garnishing wages. Besides freezing a piddling amount of overseas banking account balances, foreign holders of US bonds can’t do anything with those bonds if the US refuses to pay. They can threaten to stop doing business with the US in the future, but there’s no collection mechanism for past debts. Very different from private debts. Foreign debt holders must play nice to be paid back.

Much was made in the 80s about how Japan could sink the US by dumping bonds. Now it’s China. But think of it from China’s perspective. If it tries to dump US bonds, it would lose hundreds of billions of dollars out of the $1 trillion in bonds they hold, tanking its own economy like a reverse stimulus package. It would also drive the US dollar down, which is exactly what the Americans – both conservatives and liberals – have been hounding China to do. This would add to the Chinese collapse by stopping exports and shutting down all those Chinese factories humming along for Wal-Mart and Nike and Sears.

The effect of public debts on interest rates, inflation and expectations is complex, and borrowed funds can be used well or squandered. But, for better or worse, public debts are quite different from private debts and we should not judge them in the same light.

Saturday, February 14, 2009

Tax or Spend

There is a reason Reagan called tax cuts "Supply Side" economics, and that reason undercuts the rationale for tax cuts as a stimulus during a demand recession. There are many pros and cons of tax cuts versus government spending, but here's the economics angle.

The theory behind tax cuts is that taxes reduce the incentive to work and start businesses, because taxes reduce the reward a worker or investor earns from an increase in effort. That's why advocates of tax cuts are so wedded to tax cuts for the rich. Rich people own the businesses and create the most jobs. If we reduce the "penalty" on higher profits (i.e., taxes), rich people will want to earn more profits and grow their businesses. Workers will also want to work harder and longer hours if they take home more of their salary, but the primary way workers benefit under supply-side tax cuts is through rich people growing their businesses and profits from those businesses "trickling down" to workers.

I'll share the math below, but here's a summary of the problem with tax cuts. In a demand recession, the problem is not that workers don't want to work or businesses aren't big enough. The capacity to produce (i.e., supply) is fine. Too fine. The problem is that nobody is buying. If you cut GM's taxes, it's still not going to build a million cars that no one will buy. If you cut our unemployed friends' taxes, that still doesn't help them find jobs.

That's why even supply-side mainstream economists call for fiscal stimulus (government spending) in the current crisis. When the government builds a bridge, it employs people. It buys concrete and steel. It creates demand. If the government bought a million GM cars (just 1% of the bailout package), GM factories and supplier factories would reopen. Economically, it doesn't really matter much how the money is spent, as long as it is used to buy things, not just given to people. Ironically, Sarah Palin's "bridge to nowhere" would help close the demand recession just as much as a homeless shelter.

So why not just give the money to people? Doesn't a tax cut stimulate demand, also, since it puts money in shoppers' and business owners' pockets? Here's where the math comes in.

Aggregate demand is:

GNP = G + C + I + (X-M)

where

G = Government spending
C = Consumption
I = Investment
(X-M) = Net exports

When government spends $1M, G goes up by $1M and therefore so does aggregate demand. The $1M is spent on goods and services and becomes income for households and businesses. Some of the income is stashed away, but some -- for this example, 80% -- is spent on other goods and services, so C+I goes up by $800,000. That is the workers and business owners that get the $1M in turn spend $800,000 on food, gas, movie tickets, computers, painting services and so forth. The $800,000 they spend is income for the people and businesses they buy from, and 80% of that -- or $640,000 -- gets spent.

So the impact of $1M in government spending on aggregate demand is:

$1M (initial government spending)
+ $800,000 (multiplier effect, round 1)
+ $640,000 (multiplier effect, round 2)
+ $512,000 (multiplier effect, round 3)
etc.

...which, using the formula for the sum of a geometric series, all adds up to

1,000,000 / ( 1 - 80% ) = $5M

Now, let's say the government cuts taxes by $1M instead of spending $1M. Incomes go up by $1M, of which 80% gets spent, generating $800,000 in C+I. That's income for others, who then spend $640,000 to boost C+I more, etc.

So the impact of $1M in tax cuts on aggregate demand is:

$800,000 (initial boost in C and I)
+ $640,000 (multiplier effect, round 2)
+ $512,000 (multiplier effect, round 3)
etc.

...which adds up to

800,000 / ( 1 - 80% ) = $4M

That's $1M less impact on demand. In fact, it doesn't matter what the assumed propensity to spend (80% in this example) versus stashing money away is. The impact of $X in government spending is always $X greater than the impact of an $X cut in taxes.

Intuitively, it's easy to see why. When government gives out $1M in tax cuts, nothing is bought that requires workers and businesses to produce. There's just as many new cars on the dealership lots. There's just as much gravel in cement company stockyards. There are just as many carpenters waiting for the phone to ring. All the stimulus in demand starts when -- and IF -- people and businesses spend some part of the $1M.

On the other hand, since GNP = G + C + I + (X-M), when G (government spending) increases, GNP goes up directly because government spending programs buy cars and trucks and cement, and they hire carpenters and project managers.

The math can get more complicated, but the point holds.

For example, government spending can "crowd out" private investment by driving interest rates up. So G increases and I decreases, leaving C+I+G+(X-M) unchanged in the long run. But interest rates are irrelevant right now in our liquidity trap. Businesses are freezing investments because no one will buy what their investments produce -- not because interest rates are too high. Also, the "crowding out" effect is driven by deficits and is therefore the same whether the government spends more or cuts taxes.

Another wrinkle is that when incomes go up, (X-M) tends to go down. However that's a small drag that also affects tax cuts and government spending equally. At least, with government spending, the government can choose to direct the first round of demand (the +G) at domestic producers if it chooses to.

Government spending may take longer than tax cuts to ramp up. Tax cuts, though, might never even make it to the economy if people sit on the money or pay off debt

So, there are political arguments about how big government should be and whether rich people should be taxed more than others, but if economic stimulus in a demand recession is the goal, a dollar of government spending has more impact than a dollar of tax cuts.