Our pals at the Tax Foundation[1] (specifically the irreplaceable Erica York, Senior Economist and Research Director) took a look at the likely impact of the proposed Trump tariffs on the U.S. economy. Ms. York did us an enormous favor, summarizing the results from ten different forecasting entities. Those forecasts looked at various combinations of proposed tariffs:
- 10% on all imports (universal tariff)
- 10% universal + 60% on China
- Only 60% on China
- 20% universal + 60% on China
- One entity looked at an “aggressive US tariff scenario.”
The Tax Foundation’s first analysis assumed a 10% universal tariff. They concluded that real GDP growth would be lowered by 0.5% per year. With 10% universal and 60% on China, their estimate is real GDP growth falls by 0.8% per year. If China retaliates, the decrease is 1.2%. (Those who think these growth rate changes are insignificant may want to jump to the end of this article where you’ll find an explanation of why economists think small differences in economic growth rates matter quite a bit.)
Figure 1 Erica York
For you impatient readers, here’s the quick summary. Every one of these forecasts shows that tariffs will have a negative impact on real GDP growth. The forecasts included are not from crackpot pseudo-economists. Many of the organizations earn their income by providing analyses like this to businesses.[2] The forecast reductions in 2025 GDP range from 0.16% to 1.7%. Naturally, the higher the tariffs, the larger the negative impact. The Tax Foundation estimates ‑1.7% lower GDP assuming a 20% universal tariff, 60% tariff on China, and partial retaliation.[3] The American Action Forum predicts ‑0.16% assuming only a 10% universal tariff and no retaliation.
No, this will not be the shortest article in the history of this blog. Ms. York tracked down estimates from ten different sources, varying from economic research groups to financial market analysts all the way to international agencies. And I won’t summarize each forecast. Instead, I’ll present the summary table and mention a few of the lower points. I highly recommend reading Ms. York’s very detailed discussion.
Here’s the summary table. Their summary table is available as an Excel worksheet via the original article. Click here for my Excel calculations.[4]
Figure 2 Summary Table of Tariff Effects (click for larger image)
Universal Tariff of 10%
Nine of the ten organizations looked at this scenario. They were the American Action Forum (AAF), UBS Wealth Management, the Peterson Institute for International Economics (PIIE), Moody’s, Euromonitor, and the International Monetary Fund (IMF). Estimates of first‑year growth reductions range from −0.16% (AAF) to −1.04% (UBS, assumes retaliation). AAF estimates −0.31% with retaliation. The Tax Foundation, in an appropriate display of modesty, limits its estimate to a single decimal place: ‑0.5%.
PIIE and Moody’s gave annual forecasts for several years. Both gave estimates assuming retaliation. Only PIIE gave estimates without retaliation. For comparison, here’s what the estimates with retaliation look like.[5]
Figure 3 Annual growth change (click for larger image)
Figure 4 Cumulative GDP loss (click for larger image)
The above graphs show the four‑year estimates because that’s as far as Moody’s went. While Moody’s (‑3.61%) is a little more pessimistic than PIIE (‑2.87%), both show significant GDP losses.
No Universal Tariff With 60% on China
Only PIEE was brave enough to try this one. I’ll ignore their “no retaliation” forecast. With retaliation, 2025 shows a loss of 0.19%, “improving” to 0.12% in 2034. The cumulative GDP loss over ten years is 3.08%.
Universal Tariff of 10% With 60% on China
First, my comment on retaliation. A 60% tariff on China would almost certainly be met with a strong, swift response. To my surprise, three of the six forecasts have an estimate without retaliation.
EY[6] forecasts a ‑1.18% impact in 2025 and ‑2.34% in 2026. The Tax Foundation forecasts ‑0.8% without retaliation and ‑1.2% with. Other point estimates include Capital Economics (“up to ‑1.5%”), RBC (‑1.5% after 2 years), and The Budget Lab (‑0.64% with retaliation, ‑0.5% without).
Universal Tariff of 20% With 60% on China
Talk about pessimistic, I haven’t heard anyone propose this combination. Only the Tax Foundation and The Budget Lab were willing to make a prediction. The Tax Foundation assumed partial retaliation and The Budget Lab assumed full retaliation.[7] GDP growth gets hammered pretty well, with the Tax Foundation predicting ‑1.7% and The Budget Lab ‑0.95%. The Budget Lab adds an unspecified tariff on Mexico to the mix, forecasting a ‑1.43% hit to GDP.
Conclusion
Depressed? Don’t be. The tariff threat can be useful, but only if it is credible. Maintaining credibility may well become the major problem during President Trump’s second term. I feel pretty secure in predicting a bumpy ride. Buckle up.
Appendix: Small Growth Changes Make a Big Difference
Some of you may look at those growth changes and think they’re too small to make any difference. Over a few years, that’s true. But economists like to look at long-term impacts. Here I’ll use real disposable income per household between 1950 and 2023.[8]
Between 1950 and 2023 real personal disposable income per household (PDIH) increased by a factor of 3.2952. That implies an annual growth rate of 1.65%. In 1950 PDIH was $39,373.75. With 1.65% annual growth, by 2023 PDIH had grown to $129,742.38. That growth factor is a rough measure of the improvement in the standard of living.
Now let’s play some games. Suppose growth was 0.5% less, 1.15%. In that case, 2023 PDIH would have been $90,722.58. The growth factor is 2.3041. Households would have seen a 30% lower standard of living.
Suppose instead growth was 0.5% more (2.15%). The growth factor increases to 4.7250. PDIH increases to $186,040.17. Households would have been 43% better off compared to the actual results.
Small percentage changes make a big difference over long stretches of time. Here’s the summary table:
Figure 5 Changes in Growth (click for larger image)
- Disclaimer: I am a donor to the Tax Foundation. ↑
- Disclaimer 2: In the distant past I worked for one of these outfits: Data Resources, Inc. Look it up. ↑
- Retaliation means one or more countries subject to the tariff impose or increase their tariffs on US exports. ↑
- Notes on the Excel workbook. The workbook has seven tabs. The first (Moodys and PIIE annual 10%) uses data from the Tax Foundation table to calculate additional information. Moody’s report cumulative change estimates over four years (2025-2028). I converted those to annual rates by simply taking the difference between consecutive years. PIIE, by contrast, estimates data only for 2025 and 2034. First, I want to applaud their bravery for publishing a ten-year estimate. I converted those two figures into annual figures by calculating the ratio of the 2034 figure to the 2025. The resulting growth rate of the GDP reductions was 0.8656. I then recursively multiplied the growth rate for period t by the calculated growth rate. That is row 8 (PIIE). The cumulative impact is the rolling sum of the annual growth rates. I followed the same procedure for PIIE’s estimate of the impact of a 60% tariff on China with no universal tariff. If you see anything wrong with my process, please let me know and we can discuss better methodology. ↑
- PIEE gave 2025 and 2034 estimates. I interpolated the values using a constant growth multiplier of 0.8656. I calculated this as (r2034/r2025)(1/9)-1 = (-0.24%/-0.88%)(1/9)-1. To get the annual reduction for period t I multiplied the t-1 value by 0.8656. Moody’s gave the cumulative reduction in GDP. I took the difference between consecutive years. Criticism of this methodology accompanied by corrections or suggestions for improvement is always welcome. ↑
- EY is not Erica York. It is ey.com. Erica was kind enough to fill me in on this. ↑
- I don’t know what “partial retaliation” means. I recommend reading the Tax Foundation article. ↑
- Sources for the data. Personal income and GDP deflators are from the Bureau of Economic Analysis (https://apps.bea.gov/iTable/?reqid=19&step=2&isuri=1&categories=survey&_gl=1*zxtx3y*_ga*MjAxODY5NjE2NC4xNjY3MjU0NjE4*_ga_J4698JNNFT*MTczMjIzOTEwMy4yOS4xLjE3MzIyMzkxNDIuMjEuMC4w#eyJhcHBpZCI6MTksInN0ZXBzIjpbMSwyLDNdLCJkYXRhIjpbWyJjYXRlZ29yaWVzIiwiU3VydmV5Il0sWyJOSVBBX1RhYmxlX0xpc3QiLCI1OCJdXX0= and https://apps.bea.gov/iTable/?reqid=19&step=2&isuri=1&categories=survey&_gl=1*1dw8ilv*_ga*MjAxODY5NjE2NC4xNjY3MjU0NjE4*_ga_J4698JNNFT*MTczMjIzOTEwMy4yOS4xLjE3MzIyMzkyMjUuNjAuMC4w#eyJhcHBpZCI6MTksInN0ZXBzIjpbMSwyLDNdLCJkYXRhIjpbWyJjYXRlZ29yaWVzIiwiU3VydmV5Il0sWyJOSVBBX1RhYmxlX0xpc3QiLCI0Il1dfQ==). I deflated disposable personal income using the price deflator for personal consumption expenditure. The number of households (US households tab) is from Census Bureau table HH-1. (Link downloads the Excel file immediately. https://www2.census.gov/programs-surveys/demo/tables/families/time-series/households/hh1.xls ) That table is strange. First, the data is from most recent (2024) to most distant (1940). Second, there are two entries for several years. For example, the number of households in 1984 is either 85,407 or 85,290 (numbers are in thousands). The latter figure has footnote b: “Incorporates Hispanic-origin population controls.” Fortunately, my article really only uses 1950 and 2023 data. I ignored this issue. Here’s how to use my Excel workbook. The tab “US households working” sorts the data in the correct direction. The footnoted values are relegated to the last eight rows. The Calculations tab is the summary table, including the data from all three raw data tabs. I then calculate real personal disposable income (Real YD) using the personal consumption deflator (PCE deflator). Per household figures are in the column Real YD per household. The Results tab is where you’ll find the calculations that are the source of Figure 5. ↑
