Published on November 12, 2009

Sergio M. Marxuach

So said Aristotle, warning his students against making bold predictions based on scant evidence.  It is in this light that we should interpret the most recent change in the Government Development Bank’s Economic Activity Index, which showed a slight increase in its absolute value from 134.27 in August to 135.32 in September, which is equivalent to a month over month increase of 0.78%.

At first glance, this negligible increase in economic activity, as measured by the index, could be interpreted as a sign that “we are seeing the light at the end of tunnel.”  There are, however, several reasons to be cautious.  First, all indices have a measurement error, which, to be fair, can be attributed to the general fallibility of human beings.  In this case, we do not know what this index’s margin of error is.  If it is smaller than 0.78%, then the recently reported increase could be interpreted as significant, if it is equal to or larger than 0.78%, then the month over month change could be just a statistical glitch, the proverbial one swallow of the title.

Second, there could be an endogeneity problem with the index, as the independent variables—payroll employment, gasoline consumption, electricity consumption, and cement sales—may not be completely exogenous with respect to GNP, which is the dependent variable in this case.  There are statistical fixes to this problem, but since the methodology for calculating the index has not been disclosed it is impossible to determine if it has been corrected for endogenous relationships, if any, that may exist between one or more of the independent variables and dependent variable.

Third, even if we assume the index accurately measures economic activity in Puerto Rico, its recent performance does not bode well for the island.  The PowerPoint presentation prepared and released by the GDB includes the linear regression equation that describes the mathematical relationship between the index and Puerto Rico’s GNP at constant 2008 prices.  That equation is y=432.25x – 4,124.27, where x is the value of the index, and y is GNP.  If we calculate the index’s average value for fiscal year 2008-09 and plug it into the regression equation the GNP forecast for that fiscal year is $56.019 billion, which is 7.84% less than the GNP for fiscal year 2008 of $60.787 billion.

If we take the analysis one step further and calculate the average index value for the first three months of fiscal year 2009-10 and do the same calculation, we find that the implied GNP for the current fiscal year declines further to $54.127 billion, a decrease of 3.38% with respect to the previous fiscal year.  Again, with only three index data points for the current fiscal year it may be premature to use the index to forecast GNP.  Yet, we note that the contraction of 3.38% implied by the regression analysis is consistent with a recent estimate done by the Economist Intelligence Unit that foresees Puerto Rico GNP contracting by 3% during the current fiscal year.

In sum, the GDB Economic Activity Index could be a useful tool for determining the general trend of the Puerto Rican economy, but we should careful interpreting and drawing conclusions from it.