Presented by Robert Wade, London School of Economics, at the "Competitiveness and New Industrial Policy" workshop held at the Organisation for Economic Cooperation, 19 September 2013.

1. Governance & institutional
design of industrial policy (IP)
Robert H. Wade
LSE, September 2013

2. Preliminary point (1)
• Economic development is VERY difficult to
sustain. Adam Smith was wrong.
• Evidence: How many non-western countries
have become “developed” (productivity
criterion) in past 200 years?
• Answer: less than 10. They have in common:
strong external state enemies  societal glue
& disciplining incentives on state organizations

3. Preliminary point (2)
• IP is an “inner wheel”. Its effectiveness
depends on the “outer wheels”: eg exchange
rate, macro stability, relationship between
productivity growth and wage growth, etc.
• IP effectiveness constrained by the larger
strategy: eg export-led growth; or debt-led
consumption growth; or wage-led domestic
demand-led growth. (US – 95% of income
growth 2009-12 accrued to top 1%)

4. Export-led growth strategies reaching
limits
• X-led strategies encourage firms & govts to
keep wages low. Low wages limit domestic
demand growth, esp when many c’ies follow
same strategy.
• Demand in AC s likely remain low for some
years

5. Domestic demand strategies
• Therefore DC s shd give more weight to
domestic demand as engine of growth, &
boost wage growth
• IP shd focus less on boosting X capacities &
more on gearing production structures to
changing composition of domestic demand as
GDPPC rises
• IP shd support new production activities
important for domestic production networks

6. Examples of change of growth strategy
• (1) Singpore govt shifted from low-wage strategy
to higher-wage strategy in late 1970s, to phase
out labor-intensive, low-tech industries.
• By decree, govt drove up wages by about 20%
each year, across the board, from 1979-1981
• (2) Northern Europe, govts use “partnership”
model, involving 3 parties (govt, employers, ees)
meeting regularly to decide on policies in
medium- to long-term perspective.
• Neither approach is neoliberal “free market”

7. IP governance problem in middle-
income countries
• Public agencies in mid-income countries
commonly used as sources of patronage &
sinecures, & allocate gds & services
personalistically (eg jobs, public contracts)
• Staff commonly left to own devices while heads
busy themselves with personal business & pol
networking
• Therefore, all governance – including IP
governance -- operates in environment “hostile”
for effectiveness

8. Islands of excellence, pockets of
effectiveness
• There are also examples of public agencies
which operate like “islands of excellence” or
“pockets of effectiveness” (PoE)
• They are substantially more effective than the
country average in providing goods/services in
line with mandate, & across the country
• Question for IP governance: how to create &
sustain IP agencies which are PoEs?

9. Studies of agencies with PoE
characteristics:
• Several studies of PoE: eg by Mirilee Grindle,
David Leonard, Michael Roll, et al.
• Examples of PoE: (1) National Bank for Econ Devt
of Brazil (BNDES); (2) EMBRAPA (agriculture
research, Brazil); (3) National Agency for Food &
Drug Administration & Control (Nigeria); (4) State
Oil Company of Surinam; (5) S Korea’s Economic
Planning Board (1950s-70s); (6) Taiwan’s
Industrial Devt Bureau

10. Necessary conditions for PoE
• (1) Strong head of govt (or small coherent
elite), which has strong interest in particular
tasks being done effectively.
• Why strong interest? Diverse motives:
nationalist devt objectives; defence against
external enemies; international prestige of
head of govt. Pressure fr IFIs influential but
not decisive.

11. Necessary conditions (ctd)
• (2) Head of govt breaks with normal – patronage
-- appointment criteria, against a lot of
opposition. Instead, criteria emphasise technical
qualification, proven leadership, proven
incorruptibility.
• Director (CEO) comes fr outside the inner
elite, connected to it through “weak ties”. Hence
less vulnerable to “insider’s dilemma”.
• But, director not selected by formal bureaucratic
procedures. Because must have “strong tie” to
head of govt, not “weak tie”.

12. Necessary conditions (ctd)
• (3) Director must be free to appoint members to the
management teams, & select staff committed to mission
(“principled agents”). Some will be from outside political
elite networks (eg from private companies or overseas)
• Salaries/benefits probably higher than regular civil service.
• But performance not depend mainly on formal incentives.
Well-established in management studies that staff work
hard b/c see job as meaningful; intrinsic motivation >
extrinsic incentives. Helps effectiveness b/c reduces
director’s (Principal’s) costs of controlling staff (Agents).
• This puts added responsibility on director to foster
organizational identity, & staff’s personal responsibility for
mission

13. Necessary conditions (ctd)
• (4) Crucial to change internal & external expectations of
agency’s modus operandi.
• Keys are: (1) standardization of procedures (eg project
appraisals, project decisions); (2) regular evaluations of
agency performance
• (1) & (2) enable agency to by-pass the parallel system of
informal patronage
• Internally, standardization enables “uncertainty
absorption” (March & Simon). It raises staff confidence in
information they receive fr colleagues, do not have to
check for themselves.
• Externally, standardization enhances predictability for
clients, reduces incentives for bribes

14. Necessary conditions (ctd)
• (5) Bureaucratic autonomy necessary, b/c
agency will conflict with politicians, firms with
contrary interests
• Autonomy Paradox: autonomy is not
fixed, legal. Depends on political connections;
it is inherently relational
• Managers must manipulate their external
environment, use political connections

15. Taiwan’s Industrial Devt Bureau
• Taiwan has had central planning via market
allocation since 1950s. Eg Council for Econ
Planning & Development
• Less noticed, it has had industrial extension
service (IDB), whose agents visit factories
regularly.
• 2 way communication. “Nudges” firms up tech
ladder
• What happens when part of IDB covering
strategic sector “goes to sleep”? “Task forces”
established outside, compete with IDB.

16. Taiwan vs UK ethic of public service
• UK senior civil servant: “If I leave my office on
Friday afternoon confident that in the past
week I have done no harm, I am well content”
• In foyer of Taiwan’s Industrial Devt Bureau:
“The most important thing in life is to have a
goal, and the determination to reach it”.

17. REFERENCES
• Roll, Michael (ed.), forthcoming, The Politics of Public Sector Performance: Pockets of Excellence in Developing Countries, Routledge
• Grindle, Merilee, 1997, “Divergent cultures? When public organizations perform well in developing countries”, World Development,
25, 481-95
• Leonard, David, 2010, “’Pockets’ of effective agencies in weak governance structures”, Public Administration and Development, 30, 91-
101.
• Devlin, Robert & G. Moguillansky, 2011, Breeding Latin American Tigers: Operational Principles for Rehabilitating Industrial Policies,
ECLAC and World Bank
• Altenburg, Tilman, 2011, Industrial Policy in Developing Countries, DIE, Bonn.
• Wade, Robert, 2004, Governing the Market, Princeton University Press.
• UNCTAD Trade and Development Report 2013, 2013, Geneva.