Saturday, October 26, 2013

Predicting Employment - US Employment Will Not Achieve Pre-Recession Levels Until 2018

98% of annual employment in the United States in the past 30 years is explained by two factors: total US primary energy use in Btu's and crude oil prices in 2011 dollars per barrel.  To predict US employment going forward is simply a matter of forecasting total energy use and inflation adjusted petroleum prices.  I say simply almost as tongue in cheek, given uncertainties associated with predicting petroleum prices and energy consumption.  This approach does, however, begin to bracket the question.

The initial premise was to take a look at the correlation of aggregate energy consumption and employment.  The following chart shows that there is an amazing correlation between employment and energy use in the United States.  The energy consumption is based on primary energy use int he United States, expressed in quadrillion Btu's, including all forms of energy.  Certain observations can be drawn from the chart.  Over the period from around 1955 through 1971, energy use in the United States grew quite rapidly, and grew faster than employment.  The impact of the Arab oil embargo can be observed with a downdraft in aggregate energy, followed by another rise and steep drop in 1979.  Both of these drops in demand were a result of steep rises in oil prices.

Since 1981, the movements of energy use and employment moved much more in lockstep, with an almost uncanny correlation in the period of 2007 through 2011.

Another observation that appears uncanny is that both energy and employment tripled in the period from 1949 through 2011.  Energy use rose from 32 quads in 1949 to 97 quads in 2011.  Employment rose from 44 million in 1949 to 131 million in 2011.

The most troubling observation about this data is that our energy use as a nation appears to have stopped moving upward, hitting a peak of 101.3 quadrillion Btu's in 2007, with 2013 primary energy use around 97 quads, or 4% below the 2007 peak.  At a very basic level, we may not achieve pre-recession employment levels until such time that energy use has moved up by 5%.  Indeed, the trajectory of growth in primary energy consumption has stagnated since 2000, with current energy use of around 97 quads below the 2000 level of 98.8 quads.  As such, the inference is that we may have hit a new period in the US economy, with low growth in energy consumption, with a corresponding period of low growth in employment.

To further my understanding of the relationship between energy and employment, I performed a regression analysis of US employment from 1982 through 2011 expressed in millions, against primary energy consumption in quadrillion Btu's, and inflation adjusted crude oil prices expressed in constant 2011 dollars per barrel.  I arrived at an equation which explains 98% of employment levels in the United States.  I applied the equation first to historical energy use and oil price data, to see how it fit with actual employment levels in the past 30 years.  I then applied the equation to develop a forecast of employment, based on EIA's forecast primary energy use, and fixing crude oil prices at $96 per barrel in 2011 dollars.  The results of these analyses can be seen in the chart below.  

Based on EIA's forecast of primary energy use, the regression analysis tells us that employment levels in the United States will not achieve pre-recession levels until 2018.

The data also tells us that employment is going to grow by 0.4% per year over the next 20 years.  Unless something changes drastically, this means that the people of the United States will be suffering extended underemployment for decades to come.

No comments: