Soon after Trump’s tariffs were announced, it was noted the ‘reciprocal’ tariffs looked suspiciously like the US trade deficit divided by imports.

Ever the skeptic, I couldn’t believe it could be this simple, so I decided to check it out myself with the help of R and UN Comtrade data. Here’s my answer, neatly documented in R code.

If you’re interested in reproducing this analysis you’ll need to get an API key for the comtradr package.

Code block: project setup:

#load packages
library(googlesheets4)
library(tidyverse)
library(countrycode)
library(comtradr)

#set comtrade api key
set_primary_comtrade_key(ref_comtrade_api_key)

ref_gsheet_url<-"https://docs.google.com/spreadsheets/d/1AbPFX21KKfiCr8WWA6fwfPVKAjE5MUCzdWVitV0ZuWM/edit?usp=sharing"

#read data from google sheet 
dta_us_tariffs<-read_sheet(ref_gsheet_url)

#add countrycodes
dta_us_tariffs<-  dta_us_tariffs |> 
  mutate(iso3c= countrycode(sourcevar=country, origin="country.name", destination="iso3c"))

#assign _KS as as code for Kosovo as per Comtrade country code list (KSV is also used by some providers)
dta_us_tariffs[dta_us_tariffs$country=="Kosovo","iso3c"]<-"_KS"

#create reference list of focus countries
ref_focus_countries<-dta_us_tariffs$iso3c 

# get bilateral trade data for US 
dta_comtrade <- ct_get_data(
  reporter = 'USA',
  partner=dta_us_tariffs$iso3c,
  commodity_code = 'TOTAL',
  start_date = 2024,
  end_date = 2024,
  flow_direction = c('import','export')
)

# select focus variables and round to billion USD
dta_comtrade<-dta_comtrade |> 
  select(ref_year,partner_iso,flow_desc,fobvalue) |> 
  mutate(fobvalue= fobvalue/10^9) 

# Pivot dataframe by flow type
dta_comtrade<-dta_comtrade |> 
  pivot_wider(names_from=flow_desc, values_from = fobvalue)

#calculate trade balance
dta_comtrade<-dta_comtrade |> 
  mutate(trade_bal= Export- Import, 
         trade_bal_to_m_ratio= round(trade_bal/Import,2)*100)

#join comtrade data with tariff listing
dta_us_tariffs_and_trade_flows<-left_join(dta_us_tariffs,dta_comtrade, by=c('iso3c'="partner_iso"))

There’s more I’d like to do with this data, but for now I thought I’d share my code and the answer: yes. The reciprocal tariffs look almost identical to the trade balance to imports ratio.

@carlbergstrom.com‬ has a great thread (skeet?) on this on Bluesky.

Code block:

#create filtered dataframe where tariff > standard rate
#drop EU countries as individual reciprocal tariffs weren't published 
dta_plt_us_tariffs<- dta_us_tariffs_and_trade_flows |>
  filter(us_tariffs_applied >10,  
         EU == "N") 

#plot where tariff != reciprocal tariff
plt_us_tariffs<-ggplot(data=dta_plt_us_tariffs,
                       aes(y=us_claimed_reciprocal_tariff,
             x=trade_bal_to_m_ratio))+
  geom_text(aes(label =iso3c), size=2)+
  theme_minimal()+
  labs(x= "Imports / Trade Balance (%)",
       y="Announced 'Reciprocal' Tariff")

plt_us_tariffs