by user








Rethinking Development Assistance
A lecture at the:
Institute of Development Policy and Management
University of Antwerp
March 3, 2008
David Ellerman
University of California at Riverside
Development Debate: Cook’s Tour
Help to development or just poverty alleviation?
• Bill Easterly’s book (White Man’s Burden) argues
development aid not working so just try to alleviate the
condition of the poor.
• Agencies moving in this direction for a long time, e.g. World
What is development?
• Increase in people’s capabilities (transformation),
• Increase in people’s incomes (e.g., discovering oil)
• Examples: remittances from labor migration or Indian
casinos in US.
“Development” as a physical effect
Delivering infrastructure, social systems,
and institutions.
 Some things ‘can’ be delivered: stroke of
the pen reforms.
 ‘Vaccinating the children’ as the older
paradigm example (now it’s mosquito nets).
 Performance-based aid: international
service-providers ‘delivering’ development
goods and services.
Development Assistance
as Autonomy-Respecting Help
Autonomy-respecting form of assistance = “Helping
people help themselves” (old cliché)
Fundamental Conundrum of "assisted self-help":
• How can external influence by the helper make the doer more
autonomous or independent of external influence?
• Same Fund. Conundrum occurs across fields of human
endeavor: education, psychology, economics, politics…
“How can one person teach another person to think things out for
himself, since if he gives him, say, the new arithmetical thoughts,
then they are not the pupil's own thoughts; or if they are his own
thoughts, then he did not get them from his teacher?” (Ryle,
Gilbert 1967. Teaching and Training. In The Concept of
Education. Ed. R.S. Peters. London: Routl. & Kegan Paul, 112)
Same Conundrum in Education
“Here we actually come up against the basic problem of education, which in its
general form points to the question: How is education at all possible? If the end
of education is rational self-determination, i.e., a condition in which the
individual does not allow his behavior to be determined by outside influences
but judges and acts according to his own insight, the question arises: How can
we affect a person by outside influences so that he will not permit himself to be
affected by outside influences? We must resolve this paradox or abandon the
task of education.” (Nelson, Leonard 1949. Socratic Method and Critical
Philosophy. New York: Dover, 18-9)
“If we ask how the teacher-learner roles differ from those of master and slave,
the answer is that the proper aim of teaching is precisely to affect those inner
processes that…cannot in principle be made subject to external control, for they
are just, in essence, the processes germane to independence, to autonomy, to
self-control.” (Hawkins, David 2000. The Roots of Literacy. Boulder: Univ.
Press of Colorado, 44)
History of educational theories
about active learning and autonomy
Socrates: Socratic ignorance and teacher as
midwife not the father of learning;
 Neoplatonists including Augustine;
 Rousseau's Copernican revolution in
 Kantian themes about autonomy and
constructivist theory of the mind;
 John Dewey: main modern active learning
and contructivist educational theorist.
Bad and Good Metaphors for Learning
Bad Metaphors
Good Metaphors
Mind as passive mirror of
Mind as wax tablet on which
knowledge is stamped
Knowledge poured into mind
like water into a bucket
Knowledge transmitted from
teacher to student
Mind as active lamp of
Mind as a fountain out of
which knowledge comes
Knowledge grows in mind
like a seed properly nurtured
Knowledge awakened in
learner by teacher as midwife
Abrams, M.H. 1953. The Mirror and The Lamp: Romantic Theory and the
Critical Tradition. London: Oxford University Press.
Chomsky, Noam 1966. Cartesian Linguistics. New York: Harper & Row.
Stop Teaching; Foster Self-Learning
"If you teach a man anything he will never learn it”
(George Bernard Shaw)
"Fundamentally the staff manmust create a situation in
which members of [line] management can learn, rather
than one in which they are taught” (Douglas
"He who wants to teach a truth should place us in the
position to discover it ourselves” (José Ortega y Gasset)
"You don't just tell people something; you find a way to
use situations to educate them so that they can learn to
figure things out themselves” (Myles Horton)
Helping Relations: Helper-Doer
Thinkers who faced the conundrum in diff. fields:
Education: Teacher-Learner (Dewey, Freire)
Management: Manager-Worker (McGregor)
Psychology: Therapist-Patient (Carl Rogers)
Community: Organizer-Community Group (Alinsky)
Counseling: Counselor-Client (Kierkegaard)
Development: Agency-Government (Hirschman,
Much help is unhelpful: overrides or undercuts
capacity for self-help or autonomy.
Unhelpful Help #1: Social Engineering
“Overriding” form of unhelpful help.
 Mental model of Helper helicoptering over
maze supplying motivation and directions to
doers in maze.
 Helper as social engineer (usually an economist)
supplies plan, and
 Helper supplies “motivation” to follow plan.
 Doers’ own plans & motivation overridden.
 Alternative to social engineering is more
indirect approach.
Engineered motivation is false motivation
Action = behavior + motivation.
Genuine reform project = project + own
Aid-seeking project = “project” + external
motivation (to get aid).
Aid with Conditionalities (“carrots & sticks”)
supplies only external motivation.
No outside-in motivation for inside-out change:
like trying to shine a light on darkness.
Therefore, genuine projects must be found &
cannot be created by aid.
Don’t Try to Give “Motivation” to Doers-Help Doers Who Have Own-Motivation
“In these situations, the donor would set himself the task of
rewarding virtue ... where virtue appears of its own accord.” (Albert
“Find out what the people are trying to do and help them to do it
better.” (E.F. Schumacher)
“[Management’s] task is to provide an appropriate environment–one
that will permit and encourage employees to seek intrinsic rewards at
work.” (Douglas McGregor)
“When we confuse a physical with an educative result, we always
lose the chance of enlisting the person's own participating disposition
in getting the result desired, and thereby of developing within him an
intrinsic and persisting direction in the right way.” (John Dewey)
Alternative Indirect Approach to $-Aid
"The best kind of help to others, whenever possible, is indirect,
and consists in such modifications of the conditions of life, of the
general level of subsistence, as enables them independently to
help themselves." John Dewey
Helpers cannot supply own-motivation but must
"find virtue afoot"--and then supply some $-aid.
 But need judgment since money-moving
pressures push toward accepting camouflaged
aid-seeking projects.
Unhelpful Help #2: Benevolent Relief
“Undercutting” form of unhelpful help. Aid Addiction.
Charity corrupts; long-term charity corrupts long term.
Self-help undertaken to avoid negative or get positive
outcomes is undercut by benevolent aid which allows
doers to directly avoid negative or get positive results
without developing self-help capacity.
A problem in today’s “development industry” is
humanitarian relief repackaged as “development
assistance”—which often undercuts capacity
Relief should be designed to be developmental.
Development Aid as Moral Hazard
Insurance: moral hazard softens the
incentives to take normal precautions.
• Partial Cure: co-pay & deductibles help to
motivate investment to take precautions.
Development: Aid softens the incentives for
• Partial Cure: substantial matching or first stage
funding requirement (so help is only 2nd stage)
show own-motivation.
Scylla & Charybdis of Dev. Assist.
Scylla = Enlightened impulse to do social
engineering (mostly economists)
 Charybdis = Benevolent impulse to do
charitable relief.
 Pons Asinorum is the recognition that:
• Not the “What” of reform that counts but the
“How” if reform is to actually work. Trying to
"engineer" the "what" gets the "how" wrong.
• “Doing good” that undercuts self-help capacity
isn’t doing good.
Albert Hirschman on Alternative Indirect Approach
1940s-50s: Big Push Balanced Growth Plans
Alternative: Unbalanced growth driven by endogenous
linkages and pressures.
Rage to Conclude vs. Social Learning process
Helper as Reformmonger:
• Find where virtue is afoot on its own.
• Supply advice & aid to modestly help without overriding or
undercutting endogenous connections, linkages, and forces.
i.e., catalyze endogenous linkages to spread changes.
Alternative Approach to Aid
Global networks of smaller agencies in contrast to
today's large powerful agencies. Good example: Global
Development Network: http://www.gdnet.org/ or
Global Urban Observatory Network (GUONet).
Basic idea is to assist development-as-social-learning.
Money not the focus of development assistance so it
will not drive agendas of developing countries. Smaller
agencies cannot dominate but can be very selective.
Promote experiments, find where "virtue is afoot" and
help it indirectly whether or not it goes through central
government. Indeed, municipal governments are often
better targets of aid.
Finding and Scaling Up Learning
Center-Periphery or Hub-Spokes model
• Center does learning & decides policies
• Disseminates answers to periphery
• Evaluation checks implementation & impact of central policy.
Decentralized Social Learning Model
• Foster local experiments & self-learning
• Identify local successes and broker peer-to-peer cross-learning
& local reinvention of successes.
• Benchmarking between actual projects rather than impact
evaluation comparing project to counterfactual of doing
• See work by Donald Schön, Everett Rogers, or Charles Sabel.
Parallel Experimentation
Series vs. parallel experiments:
• If you “know what you’re doing” then experiment to move
along that branch in search tree—series experiments.
• If you “don’t know what you’re doing” then try several
experiments at once—parallel experiments.
Wright Stuff (biological version)
Different parallel experiments.
Semi-isolation between experiments.
Benchmarking between experiments.
Migration or cross-learning to ratchet up whole
Sewall Wright’s Shifting Balance Theory
Decentralized Social Learning (DSL) Projects
Center sponsors contest to “solve problem.”
Local entities voluntarily enter contest to give “best practice”
Entries must publicly state “theory” so others can see & learn.
Entries must agree to be judged by benchmarks.
Winners get “block grant” in addition to their required local
matching resources.
Center sponsors horizontal learning between laggards &
Benchmarks ratchet up performance stds as learning goes on.
Repeat as needed.
Attributes of DSL Projects
Center has problem; not “solution.”
Center’s solution-implementation capacity not assumed.
Decentralized or bottom up innovation.
Competitive contest; no entitlement to aid.
Center gives block grant to winners (who have already
shown motivation & local resource commitment).
Instead of conditionalities on “One Big Bet” at Center;
diversified portfolio of small bets w/o conditionalities.
Horizontal learning between all entrants (whether original
“winners” or not).
Ratcheting standards up.
Resistance to DSL model
Implies some restraint on part of Center not to
decide “the one best way” and then impose it.
 How can Center justify itself if it “doesn’t
know the answer”
 Parallel experimentation involves direct
comparison (benchmarking) between actual
 Center prefers “impact evaluations”: “Is our
project better than nothing? Yes! Therefore
continue the project.”
Impact of Internet
New possibilities as never before for strengthening
horizontal network connections, not just spokes
connected to hub.
Comparable effect of printing revolution on science
taking off in Renaissance:
• Wider dissemination of ideas and knowledge of experiments.
• Could compare (or benchmark) different ideas and
experimental results by putting them side by side.
Internet enables similar revolution today—a network of
local development agencies involved in decentralized
social learning.
Alternative to today’s dysfunctional development aid
For More "Left-Brain" Stuff:
Fly UP