“Keeping pace with technological change: The role of capabilities and dynamism”: Closing remarks
1. GFP Conference, 21 June 2019
“Keeping pace with technological change: The role of capabilities and dynamism”:
Closing remarks (12 minutes)
Alain de Serres, OECD
The purpose of the Global Forum on Productivity is to help advancing the
research on the sources and drivers of productivity and also to promote the
dissemination of research outcomes through conferences and workshops, such as
this one. The conference has been a real success in achieving this purpose.
We all know the nature of the challenge we are facing: how to make sure that the
rapid advances in digital technologies – and not only digital technologies – can
translate into sustainably higher productivity and income gains, but also how to
make sure that the gains are being shared as broadly as possible.
So far this has largely failed to happen: in fact to the contrary, the widespread
slowdown in productivity, that has more or less coincided with the emergence of
ICT, is persisting despite more recent and powerful waves of digital technologies.
The research presented at the Conference has helped to better understand both
the sources and nature of this new productivity paradox.
First, the paper presented by Chad Syverson (University of Chicago) to kick-off the
conference gave good reasons for why measured growth in TFP can fall initially –
and for quite some time -- following the adoption of a General Purpose Technology,
that is, precisely the type of technologies that is expected to generate a strong and
persistent boost in productivity.
o This is also the type of technologies that require massive complementary
investment in assets which for the most part are still not properly accounted
for in national statistics. And because of that, the effects of implementation
lags on productivity are exacerbated by measurement problems, resulting in
a J-curve pattern of productivity in response to a new technology.
o This in itself does not explain the slowdown but it is nevertheless providing
grounds for optimism as regards the future productivity pay-off of some of
the technologies that are still in their infancy.
o The paper and presentation by Diego Comin (Dartmouth College) also gave
reasons to be more optimistic about the prospect of a productivity pick-up in
the not so distant future, as a result of the gap between technology creation
and technology adoption being gradually closed down.
Second, recent work that Giuseppe Nicoletti (OECD) talked about this morning
has shown that the adoption of digital technologies does indeed support
2. productivity. But one finding is that adoption is proving to be slow and that it is mainly
the most productive firms that have benefitted from digitalisation so far.
o In some ways, this is showing up in one of the well-known stylised facts
behind the productivity slowdown, which is the widening productivity gap
between leading and lagging firms, a gap that is more pronounced in
digitally-intensive industries [chart].
o There is some evidence that the productivity gap is related to large
differences in the take-up and effective use of digital technologies across
firms as well as across countries.
o Data on adoption of various digital technologies across european data shows
that if most firms use broadband connections and websites, there remains
high scope for the adoption of more advanced technologies, including
decades-old technologies such as entreprise resource planning, let alone
more recent ones such as cloud computing and big data. The data show the
high degree of dispersion not only across countries but also across firms
within countries. [Chart]
o The dispersion across countries in the percentage of firms adopting specific
technologies is an indication that national policies do play a role in shaping
adoption rates. However, the fact that the rate of adoption of several
technologies remains relatively low even in better-performing countries
suggest that they all face common challenges in promoting the diffusion of
technologies, especially towards SMEs.
This raises the question of why some firms are doing so much better than others at
seizing opportunities to adopt digital technologies and making the most out of them.
Here one aspect that the Conference has clearly highlighted is the role and
importance of complementary investment in intangible assets. And, making
these complementary investments requires a mixture capabilities and incentives.
o Capabilities have to do with the technical skills required to use digital
technologies but also the good management and organisation skills needed
to make the necessary changes in business processes to use the technology
most efficiently.
o Incentives have to do with the market environment and the pressures that
firms face to invest and innovate has a means to remain competitive. So the
strength of competition policies as well the extent of regulatory barriers to
entry and exit play a key role. But all other factors that contribute to business
dynamism and the ease of resource reallocation also matter.
On skills, the second session this morning, and in particular the paper by Michela
Giorcelli (UCLA) provided one additional piece of empirical evidence on the
3. importance of managerial talent, in particular as a complement to technology. The
fact that the analysis could be achieved on the basis of a controlled experiment
makes the evidence on both the impact of better technologies and role of
managerial skills particularly valuable.
On market competition and the role it may have played in the productivity
slowdown, it is fair to say that if the vast majority of stylised facts for advanced
economies (mostly US and Europe) point to a consistent direction, the interpretation
and implications are still the object of discussions and debate.
o Chiara Criscuolo (OECD) presented further evidence on the trends towards
higher mark-ups, business concentration and M&A activities across many
industries in US and Europe and which have been documented in several
papers in recent years.
Perhaps one interesting finding is that the more evidence is shed on
this issue, the less difference there appears to be between the US and
Europe. Rises in mark-ups and concentration are prevailing in both
regions, albeit to somewhat different degrees, in particular in services.
o Chiara and co-authors have cautiously refrained from any firm policy
conclusion from their findings, especially as regards competition policies.
And, the reasons for such caution were very nicely and clearly exposed by
John Asker (UCLA) and Chris Edmond (University of Melbourne) in their
presentations which showed that the documented trends can be consistent
with competitive markets.
o In his conclusion, John has pointed to other sources of challenges for
competition, such as the growing trends in common ownership and
passive investors, which perhaps have implications going beyond anti-trust
policies and may raise questions about financial market regulation and
corporate governance.
o So far, there is little evidence on the impact and real significance of these
factors on competition and productivity. As an attempt to fill some of the gap,
the issues of common ownership and passive investors are ones that we
are currently investigating in our work programme at the OECD, and we hope
to have some results early next year.
On the other hand, if one believes in the importance of experimentation, for
innovation and productivity, then the trend declines in business dynamism ought
to be a real source of concerns for policymakers. Michael Brennan (Australian
Productivity Commission) reminded us about the distinction between business
dynamism and the share of young firms, one being about the reallocation of
resources towards more productive uses and the other being about disruptions.
Considering what we know about the contribution of young firms to growth and net
4. job creation, the decline in their share of total businesses is clearly worrying, in
particular when this is observed in tech sectors. [Chart]
The factors behind this decline have yet to be fully understood.
o The importance of regulatory barriers to both entry and exit has been shown
in several studies, notably using indicators of Product Market Regulation and
Insolvency Regimes that we have developed and that we up-date regularly
as part of our work at the OECD. Now, as Dirk Pilat (OECD) mentioned
yesterday, these indicators of Product Market Regulation have shown that
the more direct regulatory barriers to entry, such as administrative burden on
start-ups – the procedures and costs to start a business -- have fallen over
time.
o But less direct barriers to entry that we capture through this indicator do
remain high, especially in some sectors. For the most recent up-date, we
have added elements related to the regulation of lobbying activities and
what we find is that there is clearly scope to improve on that score in several
countries. And, we are also currently investigating more closely the role of
occupational licensing across many professions and sectors of activities
for a broad set of countries to see how they affect productivity and
employment.
o In his presentation, Steve Davis (University of Chicago) has also brought to
light one important factor behind the falling growth of young firms, at least in
the post-crisis period, and which is the role of housing market cycles, in
particular the boom and busts in house prices.
The evidence is impressive and quite persuasive and it would of
course be interesting to look for similar evidence in other countries
that have gone through similar boom-bust cycles, but which may also
have different practices with respect to mortgage credit and the use
of housing equity in funding entrepreneurial activities.
In terms of the main policy messages, I also want to focus on two areas that I think
the Conference has highlighted as important for countries to achieve a successful
digital transformation of their economies. And these are areas that I believe all
countries are struggling with.
o One is how best to help lagging firms, in particular SMEs, to overcome
barriers to adoption of digital technologies. How to help them overcoming the
barriers in terms of managerial and technical competencies? This is crucial
for improving the diffusion of digital technologies and narrow the gap in
productivity performance between leaders and laggards, which will help in
turn to reduce wage dispersion.
5. As was mentioned several times, creating a market environment that
continuously push firms out of their comfort zones is important, but
this may not be sufficient if many firms are facing genuine skills
constraints. In his presentation, Dan Mawson (UK BEIS) referred to
the presence of various market failures in the adoption of new
technologies and business processes, including the lack of
awareness among many firms about the possibilities and benefits.
It would be useful to collect information about programmes that
countries put in place to facilitate the diffusion to see what types of
lessons can be learnt.
o The second policy area that I think all governments are struggling to get right
is lifelong learning programmes. Yet, this will be crucial as highlighted by
some of the papers.
Thinking of occupations as a bundle of tasks and assessing how these
tasks are exposed to automation, as shown by the work of Erik
Brynjolfsson (MIT) is very useful to understand the need for future
training. It is also very helpful to get a better grip on the potential extent
and nature of job displacement. On this issue, the real concern may
be less the share of jobs at risk of displacement but more the lack of
basic skills among current adult population to undertake the type of
change in tasks that will be necessary. [Chart]
I want to finish my remark by thanking all participants on behalf of the OECD. The quality
of papers and presentation was truly impressive and generated very fruitful discussions. I
also want to thank again the organisers, the Australian Treasury and the Department of
Industry, Innovation and Science, for the very nice treatment and impeccable logistic. I
want to thank my OECD colleagues, in particular Peter Gal and Timmo Leideker.
Finally, this Conference was made possible thanks to the support that members of the
GFP are providing. We hope that you find this a very useful contribution to the policy
agenda on productivity issues.