President Trump’s trade war is in full tilt, much to the dismay of many economists and the detriment of producers and consumers across the globe. If the recent responses of nations like Canada and China are any indication, tensions will continue to escalate.
How should the Fed respond to the rising prices that will inevitably result from a trade war? And what can it do to ensure the real harm inflicted by the trade war doesn’t snowball into a general economic downturn?
Most economists in the monetary-disequilibrium tradition favor rules over discretion in monetary policy. When it comes to selecting the rule, they tend to argue that central banks should restrict their focus to targeting nominal rather than real variables. After all, central banks’ monopoly over the supply of bank reserves, a nominal variable, gives them much more direct influence over other nominal variables like inflation or nominal GDP in the long run than real variables like real GDP and unemployment. For this reason, many central banks over the past few decades have adopted some version of inflation targeting.
But not all versions of inflation targeting are equally equipped to deal with real shocks like trade wars. Economists who favor rules also tend to argue that the best type of monetary rule is one central banks can adhere to at all times, both good and bad. Which version of inflation targeting does the best job of maintaining monetary equilibrium and keeping output at its natural level through thick and thin?
Choosing the Right Target
Before we analyze what rule the Fed should use to guide monetary policy during a trade war, it’s worth recapping some of the debate over inflation targeting since it remains the most popular rule today.
Generally speaking, there are two approaches to inflation targeting. The first is to set a strict inflation target at some predetermined rate (say, 2 percent). This “strict” approach has the benefit of making the value of the monetary unit stable and predictable over time. But the benefit comes at a considerable cost. To wit, under a strict approach any deviation from the inflation target is ipso facto undesirable. This is true even when the deviation can be traced to changes on the real side of the economy, like when an oil shock causes the overall price level to rise or a technological advance leads to a dramatic reduction in prices of certain goods.
A strict inflation target would thus require central banks to respond to negative or positive supply shocks by either cutting the money supply (as in the case of a negative oil shock) or increasing it (as in the case of productivity-induced deflation) to achieve their rigid inflation target. Unfortunately, any attempt by central banks to offset these real-side-induced price changes would distort the money and loanable-funds markets and cause output to fluctuate more wildly. It would push the economy beyond its sustainable production possibility frontier in the case of a positive supply shock, or, in the case of a negative supply shock, exacerbate a real business cycle.
The limitations of strict inflation targeting have led many economists to instead support a “flexible” inflation target. As the name suggests, proponents of flexible targeting argue that central bankers needn’t panic if inflation rates vary so long as the swings stem from changes on the real side of the economy. Rather than increasing the money supply every time inflation falls, for instance, central banks should welcome below-target inflation, and even deflation in some cases, when it stems from positive supply shocks like firms’ discovering more efficient production methods or technological innovations like the 3D printer. They should likewise abstain from contracting the money supply to prevent any rise in inflation that results from negative supply shocks like, say, an oil shock or a natural disaster that destroys dozens of factories. In summary, in a world where productivity is constantly in flux, a flexible inflation target would do a better job of keeping output and employment at their natural levels than a strict inflation target.
All that said, is inflation the ideal nominal variable for central banks to target? One rule that readers of this blog are likely familiar with that is wholly consistent with both focusing the Fed on nominal variables and allowing for a flexible path for inflation is an NGDP-targeting rule. The reason an NGDP target is superior to an inflation target both in theory and practice is fairly straightforward. By maintaining a stable path for nominal spending in the economy (however measured), central banks stabilize the economy’s aggregate demand. And since a productivity norm is inherently baked into the cake of a nominal-spending target, fluctuations in real output would be limited to only those induced by changes in real, supply-side factors.
Many economists (myself included) have come out in recent years in favor of a nominal-spending target along the lines proposed by Scott Sumner and dozens of monetary economists throughout history precisely because it would minimize fluctuations in real output and maintain monetary equilibrium. It would therefore do a better job of preventing changes in the money supply from exerting any undue influence over prices or being an independent cause of mischief in the economy.
Trade Wars, Negative Shocks, and NGDP Targeting
What exactly does understanding the proper monetary response to productivity shocks have to do with why NGDP targeting is more desirable than inflation targeting in an era of trade wars?
For all practical purposes, the economic effects of expanding trade with foreigners are akin to discovering a more efficient technology or production method. Society becomes richer when individuals and firms specialize in what they can produce at the lowest opportunity cost, then trade with others who do the same. Trade also makes us richer by opening up foreign capital markets so that firms and workers can access financial and physical capital that will enable them to produce more output. Trade helps raise factor productivity and lower output prices. Recalling our earlier discussion, the increased productivity made possible by expanding our trading network can be thought of as a special case of a positive supply shock.
But a trade war that shrunk the volume of international trade would represent a negative shock. Factor productivity declines when firms’ ability to access foreign capital is restricted. Their costs of production also rise as they are forced to pay higher prices for foreign-made inputs. These higher per-unit costs are ultimately passed along to customers in the form of higher output prices. Producers are forced to scale back production and lay off workers in response to their higher costs of production and the fall in foreign demand. Workers see their real wages fall because restricting access to foreign capital and global supply chains reduces their productivity.
How would the Fed respond to such a negative shock under an NGDP-targeting regime? It would do whatever is necessary to stabilize the total flow of nominal spending in the economy, or aggregate demand. In the aggregate supply (AS) and aggregate demand (AD) framework, the Fed would simply keep the AD curve stable as the AS curve shifts inward following the outbreak of the trade war (figure 1). The economy would move up its unchanged AD curve to its new equilibrium, at which it intersects the new AS′curve. The overall price level (P) would rise; real output (Y) would fall. In effect, the Fed would respond to a trade war just as it would to a hurricane that destroys dozens of factories, an oil shock that affects firms’ costs of production, or any other negative productivity shock. This would enable it to minimize the negative effect of the trade war on real output without pushing the economy either inside or beyond its sustainable production possibilities frontier.
Figure 1. The Response of an NGDP Targeting Regime to a Negative Supply Shock
In contrast, how would an inflation-targeting Fed respond to a trade war? It would be forced to engage in contractionary policy to counteract the resulting rise in prices in order to keep inflation near its target. Unfortunately, as noted earlier, this policy response would intensify the already-detrimental impact of the trade restrictions on real output, raising unemployment and deepening the economic downturn. In the AS-AD framework, the Fed would have to actively contract the money supply to shift the AD curve in by enough to keep inflation at its desired rate (figure 2). The goal of keeping inflation stable therefore comes at the cost of greater output volatility, with the new equilibrium level of Y, Y′′, being even lower than Y′.
Figure 2. The Response of an Inflation Targeting Regime to a Negative Supply Shock
Just how much damage might a trade war inflict on our economy? Consider what happened in a previous trade war — namely, the one sparked by the passage of the notorious Smoot-Hawley tariffin 1930. Although Smoot-Hawley didn’t cause the Great Depression, it certainly contributed to it and the precipitous fall in international trade. Irwin (2006) estimates that “nearly a quarter of the observed 40% decline in imports” can be attributed to Smoot-Hawley. Bond et al. (2013) further estimate that “tariff protection reduced [total factor productivity] by 1.2%.” If Trump’s trade war were to escalate in similar fashion, it would translate into losses of tens of billions of dollars and thousands of jobs — not enough to amount to a replay of the Great Depression, but clearly significant harm.
Conclusion: Trade War, What Is It Good for?
The most important takeaway is that if we are stuck in a suboptimal world where tariffs and trade wars are here to stay, an NGDP-targeting regime would do a better job of minimizing fluctuations in real output than an inflation-targeting regime. In both cases, real output would fall as a result of the contraction in international trade. But in the case of NGDP targeting, that decline would be smaller, and monetary equilibrium would be maintained.
Thankfully, President Trump has assured us that “trade wars are good, and easy to win.” If depressed international activity is the president’s idea of winning, Americans might get sick of winning. But at least if the Fed adopts a nominal spending target, Americans might not get as sick.
Related Articles – Free Trade, Monetary Policy
What Arthur Burns Broke, Paul Volcker Fixed


Paul Volcker, who served as chairman of the Federal Reserve from 1979 to 1987, passed away this week at the age of 92. He is widely credited with ushering in a new era in Federal Reserve policy making, where much more attention is given to controlling inflation.
When President Carter appointed Volcker to the Fed, inflation was approaching double digits for the second time in less than a decade. Arthur Burns, who began his tenure as Fed chair in 1970 when inflation was around 4.90 percent, saw inflation rise to 11.51 percent in 1974 Q4, fall to 5.13 percent by 1976 Q4, and begin climbing again thereafter. Inflation rose from around 6.43 to 8.52 percent during G. William Miller’s brief tenure from 1978 to 1979.


Before Volcker, Fed chairs occasionally denied their ability to control inflation. Arthur Burns referred to cost-push inflation (in contrast to the demand-pull inflation caused by faster money growth). “The rules of economics are not working in quite the same way as they used to,” he told Congress in 1971.
There were dissenting views, to be sure. But, for most of the 1970s, they were coming from outside the Fed. Milton Friedman, for example, called Burns out at the December 1971 American Economic Association annual meeting. It was not cost-push inflation, Friedman claimed, but “erratic and destabilizing monetary policy [that] has largely resulted from the acceptance of erroneous economic theories.”
Volker changed that. He acknowledged that the Fed could bring down inflation and then set a course to do just that. Moreover, he did so with great resolve.
Engineering a disinflation is a costly proposition. The central bank must cut the growth rate of money to bring down inflation. However, cutting the growth rate of money also tends to fool producers into thinking there has been a decrease in the relative demand for their products. As a result, they produce fewer goods and services — which often means laying off workers — until they realize the error and adjust their prices down accordingly.
The underproduction problem can be mitigated, to some extent, by credibly announcing the policy in advance. If producers reduce their inflation expectations in line with the policy, they will not be fooled into underproducing.
But that is easier said than done. It is difficult to credibly announce a policy in normal times. Most folks just don’t pay that much attention to — or understand — monetary policy. It is harder still when the central bank has failed to live up to expectations in the past, since even those who do pay attention and understand how monetary policy works are unlikely to believe the Fed will do what it says it will. Hence, even when such measures are called for, cutting the growth rate of money virtually guarantees a recession.
Volcker’s disinflation was no exception. Real GDP growth fell from 6.51 percent in 1979 Q1 to −1.62 percent in 1980 Q3 and remained low through 1983 Q1. Unemployment shot up, from 5.7 percent in 1979 Q2 to 7.7 percent in 1980 Q3; by 1982 Q4, it had reached 10.7 percent. Home builders around the country pleaded for cheap credit by sending two-by-fours to the Marriner Eccles building in D.C.
But Volcker didn’t relent. Inflation came down and stayed down. Indeed, the public came to believe the Fed chair was willing to do whatever it takes to keep inflation low and steady. For every ounce of institutional credibility Burns had lost, Volcker gained a pound.
How To Stop the Proliferation of Municipal Bond Issues


It seems rather strange that in a putative democracy a handful of people can legally, if figuratively, reach into the pockets of their neighbors but it happens all the time all across America via municipal bond ballot measures.
The main problem is that the measures “pass” if the majority of those who actually vote are in favor, even if hardly anyone votes. That leads to abuses. We should change the rules and mandate that bond/tax measures must obtain the affirmative approval of over 50 percent of eligible voters, not just those with sufficient incentive, education, and information to vote.
In most areas of our lives, no means no in the sense that no decision defaults to no action. You do not have to actively dislike the advertisement of a stationary bike company to avoid buying one of its products, you can vote “no” by not taking steps to purchase one. Heck, you may even approve of its ad but that does not give the manufacturer the right to drop ship one of its high-tech torture machines to your house and dock your checking account in exchange.
The same goes for physical intimacy. A stranger does not get to lawfully have sex with you because you did not actively swipe left on his or her Tinder profile. And Tinder does not get to establish a Tinder profile for you because you did not explicitly tell it not to. Wells Fargo found that out the hard way (though arguably not hard enough).
The need to obtain explicit consent before taking somebody else’s money (or bodily fluids) is one of the key remaining features of liberty. Without it, life begins to look a lot like slavery or authoritarianism.
But the rules change when the compulsory monopoly we call government makes the rules. The original impetus behind municipal bond measures was the notion that voters need to explicitly accept the tax increases needed to service the bonds. No taxation without representation and all that. When most people voted, and where taxpayers and voters were roughly the same people, bond ballot measures approximated consent. (Why fifty percent is often considered the best threshold is another matter, but I will stipulate it here.)
Statewide bond measures pass about four out of five times. Local ones appear to pass at the same rate, even at the 55 percent threshold established in California. And issuers who fail to gain approval can try again year after year, unlike in corporate proxy resolutions where shareholders are banned from reintroducing resolutions that fail to garner sufficient votes. (The SEC, incidentally, wants to raise those thresholds.)
It is a minor miracle when voters in a town like Monument, Colorado repeatedly put the kibosh on bond measures because the issuer, often a school district, is a concentrated interest with the budget authority to hire consultants who appear to make scientific, objective cases for the “necessity” of the bond. Some of those consultants even conduct market research studies designed to help the issuer use words and arguments most likely to sway voters to click “yes” come election day. Opponents are typically individuals with jobs, families, and lives.
Unlike in the commercial sector, municipal bond issuers do not need to persuade people to their cause, they just need to create enough uncertainty, confusion, or complexity to induce most voters to abstain. While often rational in other contexts, inaction on bond measures often means tax increases because the denominator for passage, regardless of the threshold, is always the number of people who actually voted on the measure rather than the number of registered voters.
Issuers know that and use it to their advantage. A suburb of Sioux Falls, South Dakota recently passed a bond measure 1,085 to 129. That seems like a mandate except 14,700 people were eligible to vote on the measure, which went up for vote on 10 September, a time when most East River South Dakotans are busy settling their kids in school, hanging tree stands, and “gettin’ the beans in” (soybeans of course). In other words, only about 1 in 15 people explicitly approved of the bond measure but the outcome is somehow counted “democratic.” (I don’t live in that town, incidentally, and the measure did not raise taxes but merely did not lower them as a previous bond recently matured.)
Other issuers put their measures up in November but only in odd-numbered years, when voter turnout is even lower than during even-numbered years. Often, public discussion of bond measures is muted because debate might draw out voters, which issuers want to avoid because when turnouts are low measures can be won simply by mobilizing teachers and naive statists.
In response to those obvious problems, some have called for minor reforms, like mandating that all municipal bond measures come up for vote on regular election days in even-numbered years. While that would be an improvement, it misses the main point, that no (action) should always mean no (money or booty). In other words, passage of anything authorizing use of the coercive power of the state to take citizens’ money should require the assent of fifty percent plus of eligible voters, not those who turned out at the polls.
When I proposed this recently at a meeting of the South Dakota chapter of Americans For Prosperity, someone immediately objected “but then no bond measure would ever pass!” “Exactly,” was my response. But of course truly important bond measures would pass, after mature consideration and extensive public debate clarified the issues at stake.
Consider again Monument, which sits on the Front Range betwixt Denver and Colorado Springs. Traditionally, taxes and public spending there were low so it attracted childless singles and older couples. Recently, younger couples with children began moving in because it was relatively cheap and improvements on I-25 promise to reduce commute times to both metropoles. Once ensconced, though, those couples began demanding more and better schools, even though that would mean higher taxes and, ceteris paribus, lower real estate values via what economists call tax capitalization.
I do not live in Monument either and would not presume to tell its residents what type of community they should try to create. But I do believe that a nation that purports to be a democracy should encourage citizens to debate the merits of proposals openly and to have to gain explicit approval for taxes, not a bare majority of a few percent of eligible voters in an inconvenient, secretive ballot. Robust debates could raise awareness of charter schools or maybe signal to parents with young children that they should live elsewhere. Or maybe they would lead to even more financial support for public schools. At the very least, full public discussion might expose the exorbitant fees that many municipalities now pay to consultants and issuers. The point is that to win approval, issuers would have to make a case and not just slide in under the radar.
Yes, voters could turn out to defeat bond measures, as they sometimes do, but the burden of proof should fall on the issuer, especially when public school districts seek funding because they have, with few exceptions, failed to create the type of citizens who vote. NGOs like iCivics are trying to improve civics education but the real problem, especially when it comes to bond and tax issues, is the failure of public schools to teach the basic principles of economics and public finance.
Without that background, most people do not feel comfortable voting on complex bond issues. So, as behavioral finance theory predicts, many abstain and the issuers win.