6
capacity of schools in some developing
countries, some participants felt more
p re s c r i p t i v e app roa c he s m i gh t be
appropriate, thus local capacity must be
balanced against the desire to empower
local communities.
Paul Atherton commented that although
much research has studied specific
educational interventions, little research
has informed practical issues of project
implementation often confronted by aid
agencies such as DFID. DFID must work
with partners such as governments, NGOs,
and the private sector and would like to
know how to interact with these partners
in a way that will maximize educational outcomes. He also
pointed out the need for more research on educational systems
as a whole that considers how public and private sector players
interact dynamically.
Jishnu Das also strongly endorsed the need for examining
educational systems, explaining that much recent research,
especially in India, finds that private schools clearly outperform
public schools, that they react to reforms in public schools, and
that their performance often deteriorates once they receive
government subsidies. He raised the question of how we should
evaluate whether an educational system is working or not, and
suggested that rather than just focusing on test scores, it might
make sense to focus on the valuation of parents, who have their
children’s interest at heart and may care about school attributes
other than test scores (e.g., infrastructure, location, values, etc.).
This implies that a desirable feature of educational systems is that
they provide a wide range of choices for parents and be more
directly accountable to them. In discussing how to evaluate the
welfare impacts of different policies, participants also pointed
out the need to avoid focusing on just learning outcomes, but
to also consider broader social impacts as well as how scaled
up interventions might influence the system as a whole, such
as through market prices, government capacity, private sector
responses, etc.
Finally, several participants suggested that international
organizations and governments can support evidence-based
policy-making by supporting evaluation studies and sharing and
disseminating research findings, collecting and sharing detailed
administrative data on students and teachers, and systematically
measuring learning outcomes in their countries in a way that
facilitates international comparisons.
A videotape of the session can be found at:
.
HKUST IEMS cohosted the International
Conference on Human Resources and
Economic Development in June 2013,
which included many of the world’s
leading development economists studying
the challenges facing educational systems
in developing countries. The conference
also was supported by the Institute for
Advanced Study, the Department of
Economics, the Division of Social Science,
and the School of Humanities and Social
Science.
A highlight of the conference was a panel
discussion on how research can impact
educational outcomes in developing
countries, which featured comments by Paul Atherton of the
United Kingdom’s Department for International Development
(DFID), Jishnu Das of the World Bank, Michael Kremer of Harvard
University (an advisor to USAID) and Eric Hanushek of Stanford
University. Paul Glewwe of the University of Minnesota, who co-
organized the conference with HKUST IEMS Director Albert Park,
moderated the discussion, which elicited lively discussion among
participants.
In the past decade, an explosion of evaluation studies of
specific educational interventions in developing countries using
randomized control trials (RCTs) have provided new insights
into what works and what does not. Michael Kremer pointed
out that much research confirms that cost and distance strongly
influence access to education, and that RCTs have shown that
deworming, conditional cash transfer programs, efforts to reduce
mismatches between curriculum and student ability levels, and
use of local contract teachers can be effective across different
settings, while popular programs to reduce class size or provide
laptop computers are rarely successful. He explained how the
convincing research findings that deworming in Kenya could
have large impacts on school attendance at a very low cost
eventually convinced the government to adopt deworming
programs on a large scale, and that a number of Indian states are
now following suit.
Eric Hanushek and other conference participants expressed
concern about advocating specific interventions globally based
on a limited number of studies, because differences in local
capabilities and demands might make it hard to generalize
about which policies are most appropriate for a given locale. He
emphasized that research has taught us that just giving things
to schools (e.g., laptops, textbooks, equipment, money) is rarely
effective, but that incentives matter. He advocated a policy
approach in which local school systems are given a menu of
options along with strong incentives to achieve improvements
and to learn from their experience. However, given the limited
INTERNATIONAL CONFERENCE ON HUMAN RESOURCES AND
ECONOMIC DEVELOPMENT
(2013.6.24-26)
Prof. Paul Glewwe
