I want to take a new look at potential real GDP. I start with the graph of real GDP - potential real GDP according to official data.

In the past we see that real GDP would rise above potential at various times and then fall below potential during downturns. Do you really believe that real GDP was way above potential in 2000? I do not. Potential was higher than the CBO said.

Currently the data says that real GDP is $300 billion below potential. Ultimately the question is ... Is real GDP really $300 billion below potential? I do not think so. That is what I want to explore with some regressions.

I have written about this before. Here is the graph from before.

This old graph fro March 2013 shows just how much the CBO has adjusted the potential. The line was almost flat from its bottom back then. Now real GDP is shown to be rising much faster toward potential because potential has been repeatedly revised downward.

Anyway, my line (pink) was simply comparing effective labor share and capacity utilization. My line got very close for much of the past data. But now I want to use some regressions.

I take the official difference between real GDP and potential real GDP and regress it against 3 variables over time.

- Labor share index (LSI)
- Capacity utilization (TCU)
- Total monetary effect from the effective demand limit equation. (This variable assesses the looseness of monetary policy by comparing the effective demand Fed rate, Effective Fed funds rate and the 10-year treasury rate.) (Tot monet)

I first regress against all past data since 1967 for these variables.

SUMMARY OUTPUT | ||||||

Regression Statistics | ||||||

Multiple R | 0.84 | |||||

R Square | 0.70 | |||||

Adjusted R Square | 0.70 | |||||

Standard Error | 162.65 | |||||

Observations | 192 | |||||

ANOVA | ||||||

df | SS | MS | F | Significance F | ||

Regression | 3 | 11594852 | 3864951 | 146.10 | 6.89E-49 | |

Residual | 188 | 4973476 | 26454.66 | |||

Total | 191 | 16568328 | ||||

Coefficients | Standard Error | t Stat | P-value | Lower 95% | Upper 95% | |

Intercept | -7068.389 | 371.8997 | -19.0062 | 1.22E-45 | -7802.02 | -6334.76 |

TCU | 47.96182 | 3.812031 | 12.58169 | 9.86E-27 | 40.44196 | 55.48167 |

LSI | 29.1971 | 3.656311 | 7.985399 | 1.36E-13 | 21.98443 | 36.40977 |

tot monet | -17.01751 | 6.091689 | -2.79356 | 0.005753 | -29.0344 | -5.00066 |

P-values are all very low. Adjusted R squared shows that 70% of data is described by the variables. Here is the graph for plotting the equation of the regression.

This regression result agrees that currently real GDP is well below potential. But the regression implies that real GDP never got close to potential before the crisis. This does not hold up under scrutiny.

The main problem with this regression is that it includes the data for potential real GDP in the 1990's which adjusted too much downward in 1991 in my opinion. Then potential did not adjust enough upward after 1998.

So now I select out the years from mid-1990 to 2003 and the years since mid-2007. Then the regression results look like this.

SUMMARY OUTPUT | ||||||

Regression Statistics | ||||||

Multiple R | 0.92 | |||||

R Square | 0.85 | |||||

Adjusted R Square | 0.85 | |||||

Standard Error | 61.76 | |||||

Observations | 108 | |||||

ANOVA | ||||||

df | SS | MS | F | Significance F | ||

Regression | 3 | 2320435 | 773478.2 | 202.7782 | 2.67E-43 | |

Residual | 104 | 396698.1 | 3814.405 | |||

Total | 107 | 2717133 | ||||

Coefficients | Standard Error | t Stat | P-value | Lower 95% | Upper 95% | |

Intercept | -1739.55 | 276.8105 | -6.28426 | 7.82E-09 | -2288.47 | -1190.62 |

TCU | 40.82724 | 1.802787 | 22.64673 | 5.34E-42 | 37.25225 | 44.40224 |

LSI | -15.2318 | 2.560114 | -5.94965 | 3.66E-08 | -20.3086 | -10.155 |

tot mon | -7.71261 | 2.48409 | -3.1048 | 0.002453 | -12.6387 | -2.78657 |

P-values are all low. Adjusted R squared has now risen to 85% from 70%. Here is the graph plotting the regression.

This graph makes much more sense. It says that real GDP was very near potential before the crisis for a number of years. The lines match up very well before the 1990's. It also says that real GDP settled back down to potential before the 2001 recession.

The regression is not perfect yet shows a better probability that real GDP has been near potential since around the beginning of 2014.

One thing I see in the last graph is that recessions occurred when real GDP was at potential even after having been above potential. This seems strange since real GDP at potential should be a steady state and ideally stable. It may be that the dynamism of capitalism cannot tolerate a steady state.

## Comments

You can follow this conversation by subscribing to the comment feed for this post.