Dick Cheney may disagree, but most people think of large, structural government deficits as a bad thing. Sure, a case can be made for temporary, stimulative deficits, but "over the cycle", a government's books should be close to balanced. Right?

Maybe not. When a country's currency is held artificially strong by mercantilistic trading partners, perhaps the best countermove is for governments to invest in future tradables capacity by borrowing aggressively to purchase underpriced foreign goods.

Suppose mercantilistic nations subsidize exports to a country by keeping their currencies artificially cheap relative to that of the target country. Then, for a period of time, production of tradables in the target country becomes uncompetitive. Labor and capital are redirected to nontradable sectors of the economy, a current account deficit develops, and the domestic cost of capital is depressed by foreign central bank interventions. This state of affairs cannot be expected to persist forever, as currency intervention is costly to the intervening countries. (But it can persist for a long time, because the costs of intervention may be hidden and widely dispersed.) When the intervention ceases, the target of the currency manipulation will have to revive its tradables sector.

A rational response by the country whose currency is being propped up would be to devote the subsidy it receives (in the form of exaggerated buying power and cheap capital) to easing an expected future adjustment back into tradables production. But the difficulty of a reorientation to tradables is likely to increase with the length of time the tradable sector is kept artificially uncompetitive. So, an optimal policy would try to simultaneously maximize the total subsidy received, minimize the time over which it is received, and ensure that a sufficient portion of the subsidy is devoted to enhancing future tradables production.

One plausible response would be to try to maximize the rate of consumption of subsidized imports by domestic consumers. A sufficiently high rate of domestic consumption could achieve the first two goals: maximize the subsidy and minimize the time over which an adversary's intervention is sustainable. But there are many problems with this approach. First, consumption expenditures, taken as a whole, are unlikely to represent effective investment in future tradables capacity. Second, it is hard to see how a country could encourage consumption at levels higher than those desired by the intervening countries. Should a government start an advertising campaign encouraging the citizens to buy more of some particular foreign country's products? Finally, sufficiently high levels of expenditure might require many consumers to take on a great deal of debt, which they may be reluctant to do, or if they are not reluctant, may have future adverse consequences for the domestic economy.

A better response would be for the targeted country to borrow at the artificially depressed rates, and then invest the proceeds in some manner designed to enhance future domestic tradables production. For this to work, the targeted country would have to make real investments, especially by purchasing underpriced tradables, not merely save the proceeds as financial assets. (Think about it.) This borrowing and investing could be accomplished by either the private sector or the public sector of the targeted country. But, although the private sector might be eager to take on leverage in order to extract the subsidy of low interest rates, it is ill-equipped to invest the proceeds in future tradables capacity during a period when, for the foreseeable future, domestic tradables investment is expected to underform foreign tradable or domestic nontradable investments, dramatically. Also, a dramatic increase in private sector debt increases financial risk, both to the entities that take on leverage directly, and to the financial system as a whole, in ways that may not be desirable. Finally, as the ability to take on debt and profitably invest it is skewed towards the already wealthy, using the private sector to extract the foreign currency manipulation subsidy permits a foreign power to exacerbate domestic inequality, which may not be desirable.

The public sector, on the other hand, can borrow and spend at whatever level it calculates would best balance maximizing the current subsidy and minimizing the duration of other nations' interventions. The public sector is uniquely capable of making not-profit-maximizing investments on a large scale, and may wisely do so when such investment represents a "public good". Debt taken on by the public sector in its own currency can in the worst case be monetized. A sharp repricing of the currency spurred by monetization is a no-brainer when sufficiently large quantities of debt, public or private, are owed to foreigners. Mere consideration of aggressive, intentional deficit expansion to extract a currency manipulation subsidy would likely spook many private holders of domestic currency, increasing the cost and difficulty for currency interventions, and perhaps even ending them before a dime of extra public debt is actually assumed.

What would a program of massive government borrowing to invest in future tradables capacity actually look like? Well, it would be the mother of all pork programs. It would involve massive infrastructure spending; constructing well-appointed, transportation-linked industrial parks that private developers would not build on their own; fiber-lit India-style "campuses" for hosting tradable service organizations; increased capital spending on science; maybe a public organization devoted to retaining skills and knowledge in industries that have moved offshore, in case they need to move onshore again someday. Of course, many of these projects would amount to malinvestment and overcapacity, but still public sunk costs would provide private opportunities for manufacturers able to rent wonderful facilities dirt cheap. Retrospectively, these errors would function as, well, subsidies to domestic tradables producers. But prospectively, each project would have been undertaken as wise investments in the public interest, so no trade rules would be violated. The investment program wouldn't need to be perfect. It would have to be large enough to create costs for currency interveners, and should ensure that more of the currency intervention subsidy goes towards future domestic tradables production than would happen without the program. The very real dangers are that corruption and cronyism in government spending might transform capital investment programs into redistribution of consumption programs, or that corrupt or indisciplined public buyers would pay overmarket prices for tradable capital goods, subsidizing rather than creating costs for currency manipulators.

Many readers, I suspect, will be perplexed by the notion that government deficit spending explicitly to purchase more goods from a mercantilistic currency manipulator could be a strategy for ending that manipulation and eventually for bringing currant accounts into balance. Doesn't a government deficit contribute to a current account deficit? Isn't selling more goods exactly what currency manipulators are trying to accomplish by underpricing their currencies? Well, yes. But as any wrestler knows, sometimes you can throw an adversary off balance more effectively by moving too quickly in the direction you are being prodded to move than by putting up a well-anticipated fight.

Suppose a government were to borrow funds to buy up enormous quantities of steel, cement, rail, industrial machinery, or other merchanidise the production of which is dominated by mercantilistic currency manipulators. The purchasing government gets a good deal, as both the interest-rates it pays are below-market and the price it pays for goods is cheap due to the producers undervalued currencies. However, the massive purchases create inflationary pressures for the currency manipulators, twice. The price of the goods they sell (and use internally) is bid up by the sudden increase in demand. And the central banks of the manipulating countries have to buy up the extra inbound FX, in order to maintain their floors for the targeted currency. Buying the inbound currency requires expanding the domestic money supply, which contributes to domestic inflation. The central bank can fight inflation by raising interest rates or issuing sterilization bonds, but both strategies are costly. Also, as the price of some commodities is held high by sustained demand and limited capacity, fighting inflation implies accepting disruptive deflations in the price of other commodities. The currency manipulator can either acquiesce to the inflation (acquiescing to real appreciation), double down by trying to increase capacity, or cry uncle, give up the nominal peg, and let currency fluctuations and a spike in interest rates price the aggressively purchasing government out of the market. Increasing capacity is hard, slow, and counterproductive. (Just as the economy targeted by the currency manipulator faces a future adjustment into tradables production, the currency manipulator itself knows it will eventually need to rebalance out of a tradables-skewed economy.) The only way that a currency-manipulator can avoid taking losses to an overly aggressive buyer is to raise the price to the buyer of the goods it sells, by abandoning (in real or nominal terms) the floor it has tried to plant beneath the targeted country's currency.