In Policy, Tariffs, and Trade 1 [https://thomaslhutcheson.substack.com/p/policy-tariffs-and-trade] I discussed the effects of a uniform tariff 5% on imports as bandied by ex-President Trump, the revenue from which did not (because of offsetting tax or expenditure changes) reduce the deficit. The no-change-in-deficit assumption is partly out of realism (in a Trump administration any new tariff revenue would probably just be used to provide greater reduction in taxes on high income individuals) and partly for comparability to other analyses that simply ignore any revenue effects. [To analyze the effect of a tariff including the revenue effects one would need to compare the tariff with a VAT that raised the same revenue.]
To recap that analysis, the uniform tariff would raise the prices of imports and import competing goods relative to non-traded goods and services and lower the prices of exports relative to non-traded goods and services. I did not mention that, If there were significant differences in characteristics of the people employed in import competing, non-traded goods and services, and exporting, the changes in relative prices of goods and services could conceivably change the relative wages of people employed in different sectors. Likewise, if there were geographical concentration of exporting or import competing goods, the tariff could conceivably change regional incomes.
The reader may also notice that I did not say anything about the tariffs being “inflationary,” although that is often mentioned as a downside of tariff increases. In part this is on the general principle that changes in fiscal policy do not affect inflation; only the Fed controls inflation and deflation. For the same reason, tariff changes would have no effect on total employment, also controlled by the Fed. In addition, however, a uniform 5% tariff would probably produce small enough changes in relative prices to be absorbed by the Fed policy of keeping the Price of Consumer Expenditure PCE index rising at its target of 2% p.a. This is approximately equivalent to a 2.3% p.a. increase of the more widely followed Consumer Price Index (See [https://thomaslhutcheson.substack.com/p/inflation-and-measurment] for a discussion of measuring inflation with the two indices.) The Fed would be unlikely to feel that the changes in relative prices warranted additional over target inflation to facilitate market clearing.
Now let’s turn our attention to a non-uniform increase in tariffs. If a reader has ever examined a tariff schedule (the US has thousands of lines) they will be aware that tariffs are not uniform and neither is any change likely to be uniform. And the reason in not just that some sectors are better at persuading lawmakers to favor them over others. Rather the main reason is that in real life, unlike in the simplest models, goods use other goods as inputs. And when inputs into goods are not produced domestically, the very strong tendency is to exempt those inputs from tariffs. This tendency of tariffs on final goods to be higher than on inputs was brought to light by the Hungarian – American economist (and my mentor 😊) Bela Balassa. And the tendency has very important consequences for the effects of tariffs on production and the efficiency of that production. Balassa named this “effective protection” to focus on the difference in the “value added” (output minus purchased inputs of goods) that a process would have with and without the “protection” of the final good.
Let me digress to present a simplified example. Take a firm that produces $100 in widgets, using value added (wages, profits, interest on borrowed capital) of $50 and importing $50 of widgetparts with no tariff on either widgets or wiedgetparts. Now let’s suppose a tariff of 5% that allows the sales of widgets to rise to $105 but that widgetparts are exempt according to the tendency highlighted by Balassa. Note the effect on value added.
No Tariff 5% tariff % increase
Widgets $100 $105 5%
Widgetparts $ 50 $ 50 0%
Value added $ 50 $ 55 10%
In this case we would say that that the “nominal protection” (of the final output) was 5% but the “effective protection” (of the value added) was 10%[i]
The importance of the differences in effective protection is that value added is the resources that can be reallocated from one sector to another. A sector whose effective protection has increased can expand at the expense of one with static or falling effective protection and the attraction ought to be related to the differences in effective, not nominal protection. This is especially important, because it is especially overlooked, in considering resource flows out of exports and toward import substituting goods. It is through flows of “value added” that a tariff becomes a tax on exports.
Thus, even small increases in nominal protection can have large effects on resource flows when traded zero-tariff inputs are a large proportion of total cost. And much more now than in Bela Balassa’s day, domestic production has a much longer and more complex international supply chain. Therefore, it may not be so safe to pass over relative wages and regional income effects mentioned above even for small changes in nominal protection.
[Standard bleg: Although my style is know-it-all-ism, I do sometime entertain the thought that, here and there, I might be mistaken on some minor detail. I would welcome comments on these views.]
[i] I’ll stop here although it would be fun for me to go into what “effective protection” means and how to measure it when some of the purchased inputs are not traded wigdetparts, but nontraded goods and services like utilities, land rent, security, legal services. Fun because I developed my own method for taking account of these (the sector level counterpart of the exchange rate effect that increases the prices of non-traded goods in relation to exports discussed above) in my PhD thesis in 1971, before the development of “computable general equilibrium” models made my method obsolete. 😊
"The reader may also notice that I did not say anything about the tariffs being “inflationary,” although that is often mentioned as a downside of tariff increases. In part this is on the general principle that changes in fiscal policy do not affect inflation; only the Fed controls inflation and deflation. For the same reason, tariff changes would have no effect on total employment, also controlled by the Fed."
On the general principle that changes in fiscal policy don't affect inflation or no effect on total employment, could the argument about tariffs being inflationary or having effects on employment still be valid in the sense that 1) the Federal Reserve's policy could be off 2) that it still has nominal distortions, despite no real effect? You know more than me on this so I'm just curious.