REVIEW 4 major objections 6 minor 12 references
An Empirical Analysis of Tiff's Impact on American Business Formation
T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read US tariffs on China are positively associated with new US business applications from 2018 to 2025, but Chinese retaliatory tariffs carry a larger negative coefficient, largely cancelling the gain.
desk verdict Honest, simple correlation study whose main offset claim falls apart on its own stated standard of significance. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The carrying mechanism is a set of linear regressions of the form: number of business applications equals a constant plus a coefficient on US tariffs on China, plus controls, with an optional additional term for China's tariffs on the US, estimated on 88 monthly observations from 2018 to 2025. Tariff rates are converted to monthly values by weighting each rate by the number of days it was in force. The argument's pivot is the comparison of coefficient magnitudes: the US tariff effect against the retaliation effect, where the absolute value of the Chinese coefficient exceeding the US coefficient is taken as evidence that retaliation offsets the tariff's stimulus. Controls include unemployment, inflation, the federal funds rate, and quarterly government spending divided evenly into months.
What would settle it
Add a monthly time trend, seasonal dummies, or year fixed effects to the model and check the US tariff coefficient; if it becomes insignificant or negative, the claimed positive association is a trend artifact. Alternatively, test whether the sum of the US and Chinese tariff coefficients is statistically distinguishable from zero to directly evaluate the largely-offset claim.
Extended reading notes
Core claim
The paper's central claim is that the US tariffs on China during the 2018-2025 trade war had a positive effect on the return of American firms when measured by the number of new business applications, and that this effect was substantially offset by Chinese retaliation. In the specification without Chinese tariffs, a one percentage point increase in US tariffs on China is associated with 1,819 more applications, significant at the 1% level; in the specification with Chinese tariffs, the US coefficient becomes 5,142 (significant at 10%) and China's tariff coefficient is -5,296 (not statistically significant at conventional levels). The paper uses the comparison of absolute coefficient sizes to argue that retaliatory measures largely offset the benefits of protectionist policies, producing a trade-war outcome with limited net gains. It frames this as partially contradicting wholly negative earlier findings and as confirming the lose-lose theory of trade wars.
Load-bearing premise
The causal reading assumes the controls absorb every omitted factor that moved US business applications over 2018-2025, so the positive coefficient on US tariffs is not just two upward trends coinciding; the paper offers no trend, seasonality, or structural-break adjustment.
Editorial extensions
If this is right
- The positive US tariff coefficient implies that, if causal, tariff hikes during 2018-2025 were associated with a measurable increase in new business applications, giving partial support to the return-of-American-firms promise.
- Because the Chinese retaliation coefficient is larger in absolute value than the US tariff coefficient, the paper's model implies the net effect of the trade war on US business formation was close to zero or negative.
- The significant positive coefficients on inflation, the federal funds rate, and government spending imply these macroeconomic conditions also track business formation, so tariff effects are estimated only after those channels are held fixed.
- For policy, the paper's conclusion is that unilateral tariffs without diplomatic coordination are unlikely to deliver lasting net benefits, since trading partners can neutralise the gains.
Reading between the lines
- The paper does not include a time trend or seasonal terms; since business applications and tariffs both rose over parts of 2018-2025, part of the positive coefficient may reflect common upward drift rather than a causal tariff effect.
- The offset conclusion rests on comparing coefficients from two different specifications; a direct test of whether the sum of the US and Chinese tariff coefficients is zero, or significantly negative, would be a sturdier check of the largely-offset claim.
- The data are national monthly aggregates, so the paper cannot identify whether new applications appeared in manufacturing-intensive counties or simply in the broader economy; county-level panel estimation would be the natural next step.
- Business applications are not jobs or output; the paper itself notes employment, wages, and manufacturing output are left for future work, so the revitalization claim should be read narrowly.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper regresses monthly U.S. business application counts (January 2018–April 2025, 88 observations) on U.S. tariffs on Chinese goods, Chinese retaliatory tariffs, and macro controls (unemployment, inflation, federal funds rate, government spending). In the baseline specification the U.S. tariff coefficient is positive and significant (1819, p=0.009); once the Chinese tariff variable is added, the U.S. coefficient rises to 5142 (p=0.092) and the Chinese coefficient is -5296 (p=0.262). The paper interprets the larger absolute magnitude of the insignificant Chinese coefficient as evidence that retaliation largely offset U.S. tariff benefits, concluding that unilateral tariffs have limited net effects. The abstract and introduction describe the data as county-level, but no county-level variation enters the regression.
Significance. Conditional on the estimates being credible, the paper would contribute to public debate by suggesting a domestic business-formation benefit from the 2018–2025 tariffs while also quantifying an offset from retaliation. That finding would be policy-relevant and would complement the existing trade-war literature. The paper is transparent about using a simple linear model and about the non-predictive role of coefficients, and it uses public data sources; however, the current empirical design is too fragile to support the central qualitative conclusions, so the significance remains prospective.
major comments (4)
- [§3, Table 1 column 4] The central offset claim compares raw coefficient magnitudes: 5142 for U.S. tariffs versus -5296 for Chinese tariffs, and states that the latter 'substantially exceeds' the former. This comparison is not statistically supported: the Chinese coefficient has p=0.262 and the U.S. coefficient has p=0.092, and no Wald test of equality or of the sum of coefficients is reported. The paper itself states in §2 that 'the sign and significance level would be the most important term,' so using an insignificant point estimate as a precise offset measure contradicts the paper's own standard. The claim that retaliation 'largely offset' the benefits therefore rests on a statistically indistinguishable difference and should either be replaced by a formal test or withdrawn.
- [§2 model specification] The regression is a levels-on-levels OLS on 88 monthly observations with no linear or quadratic time trend, no month-of-year fixed effects, no lags, and no structural-break controls for the 2020–2021 COVID episode. Business applications are strongly seasonal and were massively disrupted during the pandemic, while U.S. tariff rates move in stepwise fashion over the sample. Under these conditions, the significant positive coefficient of 1819 in Table 1 column 1 may simply reflect two trending or seasonal series, so the paper's causal language ('positive effect,' 'return of American firms') is not justified. The authors should add trends and seasonal dummies, test for structural breaks, and report robustness excluding the pandemic period.
- [§2–§3 multicollinearity] The specification with both tariff series (column 4) is likely to suffer from severe multicollinearity, because U.S. and Chinese tariff rates move together, a point the text acknowledges in §3. The large standard errors—3,019 for the U.S. coefficient and 4,687 for the Chinese coefficient—and the instability of the U.S. coefficient across columns (1,819 to 5,142) are consistent with this problem. The paper should report pairwise correlations and variance inflation factors, or switch to first differences or event-study-type specifications, before interpreting individual coefficients in column 4.
- [§2 and abstract data description] The abstract and introduction say the analysis uses county-level business application data, but the regression uses 88 monthly national observations; no county fixed effects or county-level covariates appear in the model. This discrepancy should be corrected, and if county-level data are available, a panel specification with county and month fixed effects would provide stronger evidence than the aggregate time series.
minor comments (6)
- [§2] The construction of the tariff variable as a 'weighted sum of the tariff by the number of continuing days every month' is unclear; a time-weighted average of the tariff rate would be more natural, and the current wording leaves the units ambiguous.
- [§2] Dividing quarterly government spending by three assigns within-quarter variation uniformly, which is an arbitrary interpolation; the authors should note this or use quarterly dummies.
- [References] Some citations (Bown 2022a, 2022b) are policy briefs rather than peer-reviewed articles; please mark them as such or cite the journal version if available.
- [Title] The phrase 'TIFF’s Impact' appears to be a typo for 'Trump’s impact' or 'tariff’s impact'; please correct.
- [Data availability] The paper does not provide a replication file or data-construction details, despite using public data; adding code and data availability would improve reproducibility.
- [§4] The final policy statement that 'trade wars create lose-lose scenarios' goes beyond the empirical results, which only concern business applications; please temper the conclusion to match the evidence.
Circularity Check
No circularity: the paper estimates a direct empirical regression; the central claims are interpretive readings of estimated coefficients, not consequences of definitions or self-citation.
full rationale
The paper's derivation chain is a single reduced-form OLS regression of monthly U.S. business applications on U.S. tariffs on China, Chinese retaliatory tariffs, and controls (Section 2). The dependent variable and tariff regressors are separately measured data sources (Census Bureau, PIIE, FRED), and the coefficient gamma is estimated, not imposed. The positive-tariff claim is a reading of the estimated coefficient in Table 1, and the offset claim compares the magnitudes of estimated coefficients across specifications; neither claim is equivalent to an input by construction. No parameter is fitted to the target outcome and then 'predicted'; no result is imported from a self-citation chain; the conclusion that retaliation offsets US tariff gains is an economic interpretation of two point estimates, not a definitional identity. Statistical concerns such as omitted time trends, seasonality, or the insignificance of the Chinese-tariff coefficient are identification/robustness issues, not circularity. The analysis is therefore self-contained as an empirical association study, regardless of whether its causal interpretation is credible.
Assumptions & free parameters
free parameters (2)
- Tariff month weighting scheme =
weighted by fraction of days active
- Government spending monthly allocation =
quarterly value / 3
assumptions (4)
- domain assumption OLS exogeneity: tariff rates are uncorrelated with omitted factors affecting business applications
- domain assumption Business application counts are a valid proxy for economic revitalization promised by tariffs
- ad hoc to paper Coefficient magnitudes can be compared across specifications and variables to infer offsetting effects
- domain assumption Time series are stationary enough for OLS without trend or seasonal controls
Cite this review
Pith. "Pith review of An Empirical Analysis of Tiff's Impact on American Business Formation." pith.science (2026). https://pith.science/paper/MV5OPRUQ
@misc{pith2026250600999,
author = {Pith},
title = {Pith review of: An Empirical Analysis of Tiff's Impact on American Business Formation},
year = {2026},
howpublished = {\url{https://pith.science/paper/MV5OPRUQ}},
note = {Machine review of arXiv:2506.00999}
}
read the original abstract
This study examines whether the tariff policies delivered on promises to revitalize American manufacturing and create jobs. Using county-level business application data from 2018-2025, we analyze the relationship between tariff implementation and new business formation through linear regression analysis. Our findings reveal a statistically significant positive association between US tariffs on China and American business applications. However, when Chinese retaliatory tariffs are included in the analysis, their negative coefficient substantially exceeds the positive US tariff effect, suggesting that retaliatory measures largely offset the benefits of protectionist policies. Control variables including inflation rate, federal funds rate, and government spending show significant positive effects on business formation. These results indicate that while protectionist trade policies may stimulate domestic business formation, their effectiveness is significantly diminished by retaliatory responses from trading partners. The study provides evidence that unilateral tariff measures without diplomatic coordination produce limited net benefits, confirming that trade wars create scenarios where potential gains are neutralized by counteractions.
Reference graph
Works this paper leans on
-
[1]
Trump tariffs won't entice companies to expand us manufacturing, economic experts warn
ABC News (2025, April). Trump tariffs won't entice companies to expand us manufacturing, economic experts warn. https://abcnews.go.com/Politics/trump-tariffs-entice-companies-expand-us-manufacturing-economic/story?id=120635951 (accessed May 18, 2025)
work page 2025
-
[2]
Amiti, M., S. J. Redding, and D. Weinstein (2019). The impact of the 2018 tariffs on prices and welfare. Journal of Economic Perspectives\/ 33\/ (4), 187--210
work page 2019
-
[3]
Amiti, M., S. J. Redding, and D. E. Weinstein (2020, May). Who's paying for the us tariffs? a longer-term perspective. AEA Papers and Proceedings\/ 110 , 541–46
work page 2020
-
[4]
Bown, C. P. (2022a, July). China bought none of the extra \ 200 billion of US exports in Trump's trade deal. Peterson Institute for International Economics\/ . Last updated: December 3, 2024
work page 2024
-
[5]
Bown, C. P. (2022b, October). Four years into the trade war, are the US and China decoupling? Peterson Institute for International Economics\/ . Last updated: November 7, 2024
work page 2024
-
[6]
Cui, C. and L. S.-Z. Li (2021). The effect of the us–china trade war on chinese new firm entry. Economics Letters\/ 203 , 109846
work page 2021
-
[7]
Fajgelbaum, P. D., P. K. Goldberg, P. J. Kennedy, and A. K. Khandelwal (2020). The return to protectionism. The Quarterly Journal of Economics\/ 135\/ (1), 1--55
work page 2020
-
[8]
Federal Reserve Bank of St. Louis (2025). Federal reserve economic data. https://fred.stlouisfed.org/ (accessed May 18, 2025)
work page 2025
Show all 12 references
-
[9]
Flaaen, A. and J. R. Pierce (2019, December). Disentangling the effects of the 2018-2019 tariffs on a globally connected u.s. manufacturing sector. Finance and Economics Discussion Series 2019-86, Board of Governors of the Federal Reserve System
2019
-
[10]
Us-china trade war tariffs: An up-to-date chart
Peterson Institute for International Economics (2019, August). Us-china trade war tariffs: An up-to-date chart. https://www.piie.com/research/piie-charts/2019/us-china-trade-war-tariffs-date-chart (accessed May 18, 2025)
2019
-
[11]
Census Bureau (2025)
U.S. Census Bureau (2025). Business formation statistics. https://www.census.gov/econ/currentdata/?programCode=BFS&startYear=2018&endYear=2025&categories[]=TOTAL&dataType=BA_BA&geoLevel=US&adjusted=0¬Adjusted=1&errorData=0#table-results (accessed May 18, 2025)
2025
-
[12]
Waugh, M. E. (2019, October). The consumption response to trade shocks: Evidence from the us-china trade war. Working Paper 26353, National Bureau of Economic Research
2019
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.