Trade Deficit Model

Below are the results of running a linear regression model on various economic data (obtained from the Bureau of Economic Analysis [BEA] government site) from 1947 to the present solving for the U.S. trade deficit economic variable:

n 64

R2 0.98
Adjusted R2 0.98
SE 33.95

Term Coefficient 95% CI SE t statistic DF p
Intercept -227.5 -377.7 to -77.2 74.80 -3.04 50 0.0037
Population 0.9905 0.1843 to 1.7967 0.40138 2.47 50 0.0171
Unemployment 12.56 1.38 to 23.75 5.569 2.26 50 0.0285
Inflation 2.745 -0.866 to 6.355 1.7976 1.53 50 0.1331
GDP 0.1939 0.0017 to 0.3861 0.09570 2.03 50 0.0481
Debt 0.1539 0.0876 to 0.2201 0.03299 4.66 50 <0.0001
Tax Receipts 0.1222 -0.2344 to 0.4788 0.17755 0.69 50 0.4945
Gov Spending -0.1355 -0.5305 to 0.2596 0.19669 -0.69 50 0.4941
Budget 0.4444 0.0062 to 0.8826 0.21816 2.04 50 0.0469
Consumer Spending -1.338 -1.595 to -1.082 0.1278 -10.47 50 <0.0001
State Deficit -0.8165 -1.8917 to 0.2587 0.53530 -1.53 50 0.1335
State Social Payment 4.607 2.484 to 6.730 1.0572 4.36 50 <0.0001
Gov Social Benefits -1.488 -2.259 to -0.717 0.3838 -3.88 50 0.0003
Personal Income 0.6272 0.3585 to 0.8959 0.13378 4.69 50 <0.0001


The economic parameters used to model the U.S. trade deficit over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the U.S. Gross Domestic Product (GDP), federal government debt, federal government tax receipts, federal government spending, the federal government budget levels, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of the U.S. trade deficit (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.

The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (Trade Deficit in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (Trade Deficit in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (Trade Deficit in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (Trade Deficit in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on the U.S. trade deficit?

Increased U.S. population, unemployment, inflation, GDP, federal debt, federal budget levels, state social payments, and personal income have negative effects on the trade deficit. Consumer spending and government social benefit payments have the biggest impact to lower our trade deficit. As people and the government spend less money on exports – the trade deficit will decline.

My Book: Is America Dying? (Amazon.com, Barnes and Noble)