EDUCATIONAL OUTCOMES IN OECD COUNTRIES
31
Remember that in the baseline specification, the total value of the reform in the OECD
amounts to $275 trillion, or 645 percent of the current GDP of OECD countries.
The baseline model is based on the coefficient for education outcomes in the most
general OECD growth regression framework, i.e., 1.864 percent of additional average
annual growth for a one standard deviation increase in test scores (see column (2) of
Table 2). However, not least because of the limited sample size of the growth
regressions, the precise estimate of the growth effect of a test score increase is an open
issue. Therefore, to see how sensitive the simulations are to alternative parameter
assumptions for the growth coefficient, the first four columns of Table 9 perform the
same projection calculations for the lowest and highest estimated parameters for
cognitive skills in the different specifications of Table 2 - i.e., 1.398 and 1.968,
respectively. This leads to an estimate of the total discounted value of the education
reform for the OECD of $196 trillion and $295 trillion, respectively (or 459% and 690%,
respectively, of current GDP). Thus, while these alternative parameter choices clearly
make a difference, the bottom line of the projection analysis still is the same: No matter
which of the coefficient estimates is used, the aggregate value of the education reform is
enormous.
An alternative way to account for the imprecision of the growth coefficient estimate,
reported in columns (5)-(8) of Table 9, is to use the lower and upper bounds,
respectively, of the 95 percent confidence interval around the baseline growth coefficient
in the underlying growth regression. Results based on this analysis suggest that, with 95
percent confidence, the net present value of the education reform is between $164 trillion
and $406 trillion. While these differences are economically significant, the ultimate
conclusion remains: Even at the bottom end of the confidence interval, the value of the
reform still dwarfs any of the current stimulus packages by several orders of magnitude.
Results so far incorporate any returns that accrue by the year 2090, but disregard any
returns beyond that time horizon. Columns (9)-(12) of Table 9 report the net present
value to which the reform results aggregate for time horizons to the years 2050 and
2150, respectively. As already indicated by Figure 4, this clearly makes a huge
difference. When restricting the time horizon to 2050, the value of the reform is “only”
$36 trillion, and such a horizon neglects much of the very huge long-run growth effects.
This indicates how important it is to adopt a long-term horizon in order to see the full
effects of any education reform. Still, even by 2050, the net present value of the reform
already accumulates to 85 percent of the current GDP of the OECD countries. By
contrast, when adopting a time horizon until 2150, the (appropriately discounted) value
of the reform sums to a staggering $948 trillion, or more than 20 time the current GDP.
The baseline scenarios assume that it takes 20 years for the education reform to be
fully implemented. The first four columns of Table 10 alternatively assume reform
durations of 10 and 30 years, respectively. The faster reform implementation leads to an
increase of the reform value to $341 trillion, whereas the reform value is “only” $223
trillion if the reform takes 30 years to implement. Thus, while faster reform efforts
obviously lead to substantially higher returns, even a slow (but successful) reform, if
begun today, would have enormous impact.