The GDP Paradox --
Real?... Or Off The Mark?
Economics Without The B.S.**:
[** Double entendre intended.]
Our rate of productivity growth, as measured by GDP, has been on a declining trendline since the 1960s, our best peacetime economic growth in our 200+ years of history. And it has gotten worse since 2000, during our information technology age. Some of our "experts" say we do not have a productivity growth problem but a measurement problem by relying on GDP, an indiciator left over from our industrial economy age when we are now in a new information technology age.
I disagree
that the GDP paradox is a measurement problem. I am not denying that
there is a measurement issue, just that it is a side issue to the overall
problem of lackluster productivity gains. I would normally make an
argument that in addition to the GDP rate of increase showing a negatively
sloping trendline over the past several decades, you also have downward sloping
trendlines for expansion of industrial capacity, investment in plant &
equipment, R&D, education, labor productivity, multifactor productivity, all in declining trendlines for a number of
years, even before the Great Recession. I thought Hal Varian had good
explanations for why these trends exist, and not denying that they were trends,
and added interesting comments about capital formation for the newer
technologies versus more traditional enterprises. So let me skip
those issues.
I also thought Hal Varian had a good discussion of the
diffusion problem – with the diffusion of the technology advantage only slowly
rippling its way through our economy, although Varian sees this improving as we
go further. An excellent
point was made about optimizing the use of the newer technologies, “we have to figure
out a way to get the BEST use of it”; with a mention about Gordon's thinking we
have reached a point where we will no longer get a productivity bump out of the
information technology. Varian’s reply that
we are not there yet and what we have is “failure of imagination” I thought was
right on the mark. This diffusion
problem shows up not just in a sector analysis of our economy, but you can also
see it geographically played out with only a few metro areas (usually around 25
out of almost 400 metro areas nationally) that have high GDP productivity
(about half of the nation’s GDP) while much of the rest of the nation suffers
from below average GDP performance. And
my point regarding diffusion would be here is a technology, information
technology, that lends itself to being geographically distributed (especially
since much of it is centered in the service sector), in contrast with the
manufacturing environment (with hubs and long logistic supply chains) and yet
we see info tech centers of excellence highly concentrated in just a few
geographic areas around the country. And
we are a nation with educational resources spread out around the countryside,
to include rural areas, so you just must wonder if we are taking full advantage
of the resources we have to dedicate to the issue and get everyone up to speed.
But let’s take on the GDP measurement problem more
indirectly. Let’s use measurements not
affected by all this other nonsense of what gets measured and what
doesn’t. If we are doing so good today
with productivity because of the free resources that liberate prior cost
restraints and result in newborn efficiencies from the phantom productivity, WHERE ARE THE JOBS? The 10 years of the 1960s with annual rates
of real GDP growth that are almost 3x greater than the past 17 years and a
workforce that was less than half what ours is, produced (had a net gain: jobs added
minus jobs lost) over 17 million jobs – the 17 years of Bush/Obama/and 11
months of Trump have a net gain of less than 15 million; even if you add the 6
million job openings that only gets you to a little over 20 million, still not
a good comparison with the ‘60s with half the workforce. Take 1965 with a workforce of approximately
75 million, it produced more jobs than 2016 with a workforce of almost 160
million (2.9 million vs 2.2 million); and it is likely that 2016 will be better
than 2017. And I use 1965 because that
is before the Vietnam buildup influences the economy.
This is the creative/destruction argument of the
impact of technological innovation in a nutshell. The 1960s were just as dynamic as what we
have been going through in the last 25 years.
In the 1960s we had the destruction, but we also had the creation; and,
a long sustained booming economy with the productivity gains widely shared in
society. In the last 25 years we have
had the destruction, but only a tepid creative recovery; and, the productivity
gains have resulted in a maldistribution of wealth with increased inequality in
comparison to the ‘60s. The job creation
has been a lot weaker and as Varian has said we need the job creation and
productivity boost to overcome the looming demographic shift. If we are losing the dynamism in our
society that makes for that creative/destructive energy that propels economic
growth, is there an opportunity for public policy remediation? Perhaps we can get back to using fiscal
policy more aggressively like we did in the 1960s.
Getting
the nation’s productivity picture is important for maintaining and improving
our standard of living as we go on – human progress is linked with economic
development and democratic societies function best when that progress is
broadly shared. So, by raising the GDP measurement issue for keeping
track of our progress is either a distraction or a key indicator for
consideration. Once again, I think while it is an issue, it is a
side issue to the overall problem of lackluster productivity gains and the long
trendline of contractionary economic policies that have resulted in the United
States emphasizing consumption over savings and investment that have affected
our international trade imbalances as well as our continuous grand fiscal
deficits.
