CONTEMPORARY MODELS OF DEVELOPMENT AND

UNDERDEVELOPMENT
INTRODUCTION
After more than a half century of experience with attempting to encourage
modern development, it has now been recognized that development is both
POSSIBLE & extremely DIFFICULT to achieve.

Since the late 1980s, significant strides have been made in the analysis of
economic development & underdevelopment:
In some cases, ideas of the classic theories reviewed so far have been
formalized (and in the process) their logical structure and policy
significance have been clarified & refined.
At the same time, the analysis has also led to entirely new insights into
what makes development so hard to achieve (as in Sub-Saharan Africa) but
also possible to achieve (as witnessed in East Asia).
Thus, an improved understanding of IMPEDIMENTS and CATALYSTS of
economic development is of paramount importance to broaden its reach to
countries that still remain underdeveloped.
In this chapter, we review some of the most influential new models of
economic development.
A MAJOR THEME in this new perspective is incorporating Problems of
Coordination among economic agents, such as among a group of firms. 4-2
Underdevelopment as a COORDINATION FAILURE

Many newer theories of economic development have emphasized

complementarities between several conditions necessary to get
sustainable economic development underway.
 A COORDINATION FAILURE is a state of affairs in which the inability of
agents to coordinate their behavior (choices) leads to an outcome
(equilibrium) that leaves all agents worse off than in an alternative
situation that is also an equilibrium.
It can occur even when all agents are fully informed about the
preferred alternative outcome because of difficulties of coordination.
In general, when complementarities are present, any action taken by a
firm, worker, organization or government increases the incentives for
other agents to take similar actions.
These complementarities often involve investments whose return
depends on other investments being made by other agents.
Such network effects include the BIG-PUSH MODEL (in which production
decisions by firms are mutually reinforcing), and O-RING MODEL (in
which the value of upgrading skills or quality depends on similar
upgrading by other agents).
4-3
Coordination Failure Cont’d

One example of a COMPLEMENTARITY is the presence of firms using

specialized skills and the availability of workers with those skills.
Firms won’t enter a market if workers do not possess the skills
needed, but workers won’t acquire the skills if there are no firms to
employ them: although everybody would be better off if workers
acquired the skills & firms invested, it may not be possible to get to this
better outcome solely by the market mechanism.
Another example concerns the commercialization of agriculture in
LDCs: Farmers must somehow get their products to markets while
convincing distant buyers of their quality.
Middlemen play a key role in the development of agricultural markets
by effectively vouching for the quality of the products they sell, since
they know the farmers from whom they buy as well as the product.
And since it is difficult to be an expert in the quality of many products, a
specialized agricultural market needs to emerge; and this requires a
sufficient number of concentrated producers with whom a middleman
can work effectively.
4-4
Coordination Failure Cont’d

But without available middlemen to whom the farmers can sell, they
will have little incentive to specialize in the first place and will prefer
to continue producing their staple crop or a range of goods primarily
for household consumption or sale within the village.
The result can be an underdevelopment trap in which a region remains
stuck in subsistence agriculture.
BOTTOM-LINE: Such coordination problems can leave an economy
stuck in a bad equilibrium: at a low PCI level or growth rate.
In many cases, the presence of complementarities creates a classic
“chicken & egg” problem: Which comes first, the skills or the demand
for skills? Specialization or the middlemen?
 Often the answer is that the complementary investments must come at
the same time, through coordination. This is particularly so when there
is a lag between making an investment & realizing the return on it.
In this case, even if all parties expect a change to a better equilibrium,
they will still be inclined to wait until others made their investments.
 Here, government policy can play a crucial role in coordinating joint
Investments.
4-13
Coordination Failure Cont’d

 MULTIPLE EQUILIBRIA: A Diagrammatic Approach

MULTIPLE EQUILIBRIA is a condition in which more than one equilibrium
exists: such equilibria may sometimes be ranked, in the sense that one
is preferred to another; however, the market alone will not move the
economy to the preferred outcome.
The standard diagram to illustrate multiple equilibria with possible
coordination failure is shown in Fig. 4.1: the basic idea in this S-shaped
curve is that the BENEFITS an agent receives from taking an action
depends POSITIVELY on how many other agents are expected to take
the action, or on the extent of those actions.
E.g. the price a farmer can hope to receive for his produce depends on the
number of middlemen who are active in the region, which in turn depends
on the number of other farmers who specialize in the same product.
Coordination is really not much of a problem when the group to be
coordinated is small: the parties know each other & share common
interest, and they can communicate at low cost.
But when the problem is as complex as achieving development, the
Solution to coordination problems are far more difficult. 4-14
Coordination Failure Cont’d

Equilibrium is found where the “PRIVATELY RATIONAL DECISION FUNCTION”

(i.e. the S-shaped curve) crosses the 45º line. This is because, in these
cases, agents actually observe what they expected to observe.
Suppose that firms expected no other firms to invest, but some firms
did anyway, implying a positive vertical intercept in the diagram.
But then, seeing that some firms did make investments, it would
not be reasonable to continue to expect no investment!
Firms would have to revise their expectations upward, matching their
expectations to the level of investment they actually see. But if firms
now expect this higher level of investment, firms would want to invest
even more.
This PROCESS OF ADJUSTMENT OF EXPECTATIONS would continue until
the level of Actual Investment is just EQUAL to the level of Expected
Investment: i.e. the number EXPECTED to take an action is EQUAL to the
number that ACTUALLY take that action.
At that level, firms would have no reason to adjust their expectations
any further.
4-7
Coordination Failure Cont’d

4-8
Coordination Failure Cont’d

So the notion of an EQUILIBRIUM in such cases is one in which all

participants are doing what is best for them, given what they EXPECT
others to do, which in turn matches what others are ACTUALLY doing.
In Fig. 4.1, the curve cuts the 45º line 3 times, any of which can be
an equilibrium: that is the possibility of MULTIPLE EQUILIBRIA!
Of the three, D1 & D3 are “stable” equilibria, because any slight change
in expectations to above or below these levels causes agents to adjust
their behavior that restores the original equilibrium.
A closer inspection of these 2 points reveals that the S-shaped curve
cuts the 45º line from ABOVE, which is often considered as a hallmark
of a stable equilibrium.
The equilibrium at D2 cuts the curve from below, and hence is UNSTABLE
This is because, if a few less were expected to make the decision, the
equilibrium would be D1; and if a few more, the equilibrium would
move to D3. Thus, D2 would be an equilibrium only by chance!
In practice, we think of an unstable equilibrium such as D2 as a way
of dividing ranges of expectations over which a HIGHER or LOWER
STABLE EQUILIBRUM WILL HOLD WAY.
4-17
Coordination Failure Cont’d

Generally, when jointly profitable investments may not be made without

coordination, multiple equilibria may exist, in which the same individuals
with access to the same resource & technologies could end up in either a
Good or a Bad situation.
It is the opinion of many development economists that most LDCs are
essentially caught in such circumstances economy-wide.
A classic case of this problem in economic development concerns
coordinating investment decisions where the VALUE (rate of return) of one
investment depends on the presence orextent of other investments.
All are better off with more investors or higher investment rates, but the
market may not get us there without government policy support.
Another problem illustrated by the graph in Fig. 4.1 could be that the amount
of effort each firm in a LDC exerts to increase the rate of technology
transfer depends on the efforts exerted by other firms.
Introducing new technology often has spillover effects for other firms, but
the Possibility of Multiple Equilibria reveals that the availability of a better
technology does not mean that it will be adopted.
4-10
Coordination Failure Cont’d

Market forces can generally bring us to one of the equilibria, but they are not
sufficient to ensure that the best equilibrium will be achieved; and nor do
they offer mechanism to get unstuck from a bad equilibrium and move
toward a better one.
It is important to note here that STRATEGIES for coordinating a change from
a less productive to a more productive set of mutually reinforcing
expectations can vary widely.
 For instance, changing expectations may not be sufficient if it is more
profitable for a firm to wait for others to invest rather than to be a
“pioneer” investor; and such a wait & see strategy on the part of each firm
may prevent the achievement of the desired outcome.
 In such cases, government policy (incentive schemes, investment on physical
& human capital or even a direct government involvement, besides changing
expectations) is generally needed to move the economy to a better
equilibrium.
CONCLUSION: Coordination failure is pervasive in LDCs’ economies, resulting
in Low-level (PCI, growth rate, etc) Equilibrium Trap.
The fundamental problem of coordination failure essentially arises because
economic agents keep doing inefficient things since it is rational to keep
doing them, and it remaining rational for as long as others keep doing
4-11
them.
Coordination Failure Cont’d

Coordination problems can feature themselves in many ways:

Resistance to Change (even if change would be beneficial to all)
Wait and See Attitude may predominate due to the lack of incentives
for the first mover.
Sometimes economic agents may be able to coordinate on their own to
achieve a better equilibrium. But in many cases, public policy is called
for to overcome the resulting vicious circles of underdevelopment.
There is thus a room for (costless) government policy to boost a
change in EXPECTATIONS:
Change in Norms, Laws, Policies (INSTITUTIONAL CHANGE)
Provision of guarantees to the first mover (INCENTIVES)
However, change in expectations alone may not be enough to launch
virtuous circular causation if PECUNIARY EXTERNALITIES (a positive or
negative spillover effect on an agent’s costs or revenues) are also
involved; in which case SUBSIDIES may also be necessary.
In fact, what is needed is a CONCERTED EFFORT to get development
Started i.e., A BIG PUSH. 4-12
THE BIG PUSH: Starting Economic Development

It seems that whether an economy has been growing sustainably

for some time or it has been stagnant is a BIG DEAL for subsequent
development: i.e., If growth can be sustained for a reasonable period,
it is unlikely for economic development to later get off track for
long (except for temporary setbacks caused by short-run shocks).
According to Rostow, that once economic development is underway, it
can never be stopped.
We should note here that while it is very difficult to get modern
economic growth underway in the first place, it is, however, much
easier to maintain it once a track record has been established.
But why should it be so difficult to start MODERN GROWTH?
Many development economists argue that several market failures
work to make economic development difficult to initiate, notably
pecuniary externalities (i.e., spillover effects on costs or revenues).
The most famous coordination failures model in the development
literature is that of the “BIG PUSH” pioneered by Rosenstein-Rodan,
Who first raised some of the basic coordination issues.
4-13
The Big Push Cont’d

He indicated several problems associated with initiating industrialization

in a subsistence economy. The problem is easiest to perceive if we
start with the simplifying assumption that the economy is closed.
In this case, the question is: who will buy the goods produced by the first
firm to industrialize? Starting from a subsistence economy, workers
will not obviously have the money to buy the new goods.
The first factory can sell some of its goods to its own workers, but no
one spends all income on a single good. Each time an entrepreneur
opens a factory, the workers spend some of their wages on other goods.
 So the profitability of one firm depends on whether another one
opens, which in turn depends on its own potential profitability, and
that in turn depends on the profitability of still other firms.
This circular causation is the typical pattern of coordination failure
problem that we are now familiar with.
Moreover, the first factory has to train its workers, who are accustomed
to a subsistence way of life. The cost of training puts a limit on how
high a wage the factory can pay and still remain profitable.
4-23
The Big Push Cont’d

However, once the first firm trains its workers, other entrepreneurs, not
having to recoup training costs, can offer a slightly higher wage to
attract the trained workers to their own new factories.
But the first entrepreneur, anticipating this, won’t pay for training in the
first place. No one is trained, and industrialization never gets underway!
The Big Push is a model of how the presence of market failure can
lead to a NEED for a CONCERTED Economy-wide & Policy-led EFFORTS
to get the long process of development underway and sustain it.
The BIG PUSH Graphical Model has the following 6 assumptions:
 There is only one factor of production, labor (L), that has fixed supply.
 The labor market has two segments: the traditional sector pays workers
wage W = 1, while workers get a wage W > 1 in the modern sector.
 TECHNOLOGY: there are N types of products; in the traditional sector
(with CRS) one worker produces one unit of output, while in the modern
sector, there are IRS (where no product can be produced unless a minimum
of workers are employed). And each producer in the modern sector faces
the same production function for producing any product.
4-15
The Big Push Cont’d

 Consumers spend an equal amount on each good: each good receives a

constant & equal share of expenditure out of national income Y, so that
consumers spend equal amount, Y/N, on each good (i.e., NO Savings).
 The economy is closed: no international trade.
 MARKET STRUCTURE: Perfect competition in the traditional sector (free-
entry, zero economic profits & the price of each good will be 1, equal to
the marginal cost of labor). A Monopoly in each market of the modern-
sector is assumed: i.e., at most, one modern-sector firm can enter each
market, and this is a limitation due to IRS. The monopolist charges the
same price of 1 if it decides to enter the market, since competition
from the traditional-sector producers causes the modern firm to lose all
of its business should it charge a price higher than 1.
MULTIPLE EQUILIBRIA CONDITIONS: Given these assumptions, we
now can characterize cases requiring a BIG PUSH.
We start with a traditional economy with no modern production:
Now let a potential firm with modern technology consider whether to
enter the market, which depends on whether it is profitable to do so.

4-16
The Big Push Cont’d

Given the size of the fixed cost, entry depends on 2 considerations:

1. How much more EFFICIENT is the modern sector relative to the
traditional sector, and 2. How much higher are wages in the modern
sector relative to those prevailing in the traditional sector.
In fig.4.2, the production functions are represented for the 2 - types of
firms for any industry: the traditional producers utilize a linear technique
with slope 1, with each worker producing 1 unit of output.
The modern-firm requires F workers before it can produce anything (fixed
input/cost) but after that it faces a linear technique, with slope 1/c > 1.
Product Price is 1, so revenue, PQ, can be read off the Q - axis.
For the traditional firm, the wage bill line coincides with the production
line while for the modern firm, the wage bill line has a slope W > 1.
Now, first consider the wage bill-line W1 passing below point A; with this
relatively low modern wage, revenue (1•Q1 = Q1) exceeds costs and the
modern firm will pay the fixed cost F & decides to enter the market.
This outcome is more likely if the firm has lower FIXED COST or LOWER
MARGINAL LABOR REQUIREMENTS, and if it pays relatively LOWER WAGE.

4-17
The Big Push Cont’d

c2

The labor requirements for producing

any product in the modern sector
take the form L = F + c.Q, where c < 1
is the marginal labor required for an
extra unit of output.

4-18
The Big Push Cont’d

Since symmetry is assumed for each good, so if a modern firm finds it

profitable to enter, the same incentives will be present for all firms, and
the whole economy will industrialize through market forces alone, and
demand is now high enough that we end up at point B for each product.
i.e., a COORDINATION FAILURE need not always happen; it depends on
the technology & prices (including wages) prevailing in the economy.
If a wage bill-line W2 holds, passing b/n A & B, the firm would not enter
if it were the only modern firm to do so in the economy, as it would
incur losses (since revenue Q1, is less than costs, C2).
But if modern firms enter in each of the markets, then wages are raised
to the modern wage in all markets, and income expands.
Firms in each sector can now sell all of their expanded output (at B),
produced using all of their labor allocation (L/N) b/c of sufficient demand
from workers & entrepreneurs in the other industrializing sectors.
With wage W2, point B is profitable after industrialization since it lies
above the W2 wage bill-line.
There are thus TWO-EQUILIBRIA with wage W2: one in which firms with
modern techniques enter in all markets, raising wages & output.
4-19
The Big Push Cont’d

And another in which no modern firm enters, and wages & output
remain lower. The equilibrium with higher output (B) is clearly better,
but in general the market will not get the economy there by itself.
A final possibility is at a wage line like W3, passing above point B. Even if
a modern firm entered in all sectors, all of these firms would still lose
money. Again the traditional technique would continue to be used.
In general, whenever the wage bill-line passes BELOW Point A, the
market alone is likely to lead the economy to MODERNIZATION; and
whenever ABOVE Point A, it will not. The STEEPER the modern sector
production line or the lower the fixed cost (the more efficient it is),
the more likely that the wage bill-line will pass below Point A.
Ifthe wage bill line passes above B, it makes no sense to industrialize.
But if the wage bill line passes b/n A & B, it is efficient to industrialize,
but the market will not achieve this on its own.
Here we have a problematic situation that results in TWO - EQUILIBRIA:
one in which there is industrialization & the society is better off (Point
B), and one without industrialization, (Point A). But the market will not
take us from point A to B because of a coordination failure.
4-29
The Big Push Cont’d

Thisis precisely a case in which a Big Push is Necessary: encouraging the

simultaneous expansion of the modern sector in many industries!
SUMMARY: The simple model of Big Push illustrates why economic
development has often been difficult to achieve in traditional economies,
despite the availability of modern technology. Two Key points to note here:
1. It is not necessary for all sectors to industrialize to get sufficient push for
some to do so; it is only necessary that a sufficient number industrialize so
as to generate enough national income (through the higher industrial wage
& positive profits from the industrialized sectors) to make industrialization
minimally profitable.
2. Each firm’s failure to realize the impact of their joint investments on
demand for other firms’ goods represents a very small distortion by itself;
but when added up across all of the sectors, the resulting distortion,
namely FAILURE TO INDUSTRIALIZE AT ALL, is a very large one!
The government is best suited to INDUCE THE COORDINATION (East Asian
success) since the coordination failure can not be solved by one private
“super-entrepreneur”. BUT WHY NOT?!
What is thus needed to solve the problem is PUBLIC COORDINATION OF
4-21
ACTIONS OF THE PRIVATE INVESTORS.
ADDITIONAL NOTE: ANY ONE INTERSTED CAN READ THIS PART!!!

Michael Kremer’s O-Ring Theory of Economic Development

O-ring theory of economic development is a theory of coordination

that indicates that production consists of MANY TASKS, all of which
must be successfully completed for the product to have FULL VALUE.
i.e., modern production (in contrast to traditional crafts production)
requires that many activities be done well together in order for any of
them to amount to high value.
This is a form of STRONG COMPLEMENTARITY and is a natural way of
thinking about Specialization & the Division of Labor, which along with
ECONOMIES OF SCALE is the hallmark of advanced economies in general
and modern industrial production in particular.
Michael Kremer uses the 1986 space shuttle Challenger as a metaphor
for coordinating production in his Theory of Economic Development.
The Challenger has thousands of components, but it exploded because
one relatively simple component, the O-rings, malfunctioned.
Somewhat analogously, he proposed a FUNCTION in which production
consists of many tasks (either simultaneous or sequential), all of which
must be successfully completed for the product to have full value.
4-22
Kremer’s O-Ring model Cont’d

 This function describes production processes subject to mistakes in

any of several tasks, and it does not allow the substitution of quantity.
Highly skilled workers, who make few mistakes, will be matched
together, with wages and output rising steeply with skill.
Kremer thinks the O-ring theory can explain why rich countries
specialize in more complicated products, have larger firms, and have
astoundingly higher worker productivity & average incomes than poor
countries.
Suppose that a production process is broken down into n - tasks. There
are many ways of carrying out these tasks, which for simplicity are
ordered strictly by the level of skill, q, required, where 0 ≤ q ≤ 1.
The higher the skill, the higher the probability that the task will be
“successfully completed”, which may mean that the part created in
this task will not fail.
Kremer’s concept of q is quite FLEXIBLE. Other interpretations may
include a quality index for characteristics of the good: Consumers
would be willing to pay more for higher-quality characteristics.

4-23
Kremer’s O-Ring model Cont’d

If, say, q = 0.90; then, among other interpretations, this could mean:
(a) that there is a 90% chance that the task is completed perfectly, so
the product keeps maximum value, and a 10% chance that it is
completed so poorly that it has no value; or (b) that the task is always
completed well enough that it keeps 90% of its maximum value.
The production function assumed is a simple one: Output is given by
multiplying the q values of each of the n tasks together, in turn
multiplied by a term, say, B, that depends on the characteristics of the
firm and is generally larger with a larger number of tasks.
If each firm hires only two workers, then the O-ring production function
looks like this:

We also make three other simplifying assumptions: (1) Firms are risk-
neutral, (2) labor markets are competitive, and (3) workers supply labor
in-elastically (i.e., they work regardless of the wage).
One of the most prominent features of this type of production function
is what is termed POSITIVE ASSORTATIVE MATCHING: i.e., workers with high
skills will work together and workers with low skills will work together.
4-24
Kremer’s O-Ring model Cont’d

When applied to comparing economies, this type of matching implies

that high-value products will be concentrated in countries with high-
value skills.
In this model, everyone would like to work with the more productive
workers: i.e., if efforts are multiplied, one will be more productive
when working with a more productive person.
This can be seen easily if we imagine a four-person economy, in which
there are two High-skill (qH) workers and two Low-skill (qL) workers.
The four workers can be arranged either as MATCHED SKILL PAIRS or
UNMATCHED SKILL PAIRS. Total output will always be higher under a
matching scheme because:

Since TOTAL VALUE is higher with SKILL MATCHING than with SKILL MIXING,
the firm that starts with high-productivity workers can afford to bid more
to get additional high-productivity workers, and it is profitable to do so.
If there are many classes of skill or productivity, first the highest skill
workers get together, then the next highest, and so on, such that skill
matching results as a CASCADING PROCESS.
4-25
Kremer’s O-Ring model Cont’d

The implication for the business world is that some firms and workers,
even an entire low-income economy, can fall into a trap of low skill and
low productivity, while others escape into higher productivity.
To illustrate how firms with high-skill workers could & would pay more
to get other high-skill workers or would have more incentive to upgrade
skills among existing workers, let there be 6 workers: 3 have q = 0.4
while the other 3 have q = 0.8, and are grouped together in two firms.
Now suppose that the q of one of the workers in the first firm rises from
0.4 to 0.5 (say, due to training). Similarly, let the q of one worker in the
second firm rise to 1.0 from 0.8; i.e., a 25% increase in the quality of
one worker, which leads to a 25% rise in output quality in each case.
But starting from a higher level of quality, that 25% clearly translates into
a much larger point increase for the second than for the first firm:
The 1st firm goes from (0.4)(0.4)(0.4) = 0.064 to (0.4)(0.4)(0.5) = 0.080; leading
to a point change of 0.016 (i.e., 0.080 - 0.064); and 0.016/0.064 = 0.25,
which is a 25% increase.
While the 2nd firm moves from (0.8)(0.8)(0.8) = 0.512 to (0.8)(0.8)(1.0) = 0.640
the change here is 0.128, which is again a 25% increase.
4-26
Kremer’s O-Ring model Cont’d

But the point value of the increase is much higher (8 times greater) for a
doubled point value investment: 0.2 in the 2nd firm Vs 0.1 in the 1st firm.
If a firm can increase quality in percentage terms at constant MC or a
not too quickly rising cost, there is a VIRTUOUS CIRCLE in that the more
it upgrades overall, the more value it obtains by doing so. Accordingly,
wages would increase at an increasing rate as skill is steadily raised.
 IMPLICATIONS - the analysis has several important implications:
Firms tend to employ workers with similar skills for their various tasks
Workers performing the same task earn higher wages in a high-skill
firm than in a low-skill firm. And since wages increase in q at an
increasing rate, wages will be more than proportionally higher in MDCs
than would be predicted from standard measures of skill.
The model can help explain the international BRAIN DRAIN: It is often
observed that when a worker of any given skill moves from a LDC to a
MDC, she/he receives a higher wage for using those same skills.
Workers would have a greater incentive to acquire more skills when
those around them have higher average skills.
4-27
Kremer’s O-Ring model Cont’d

O-ring effects magnify the impact of local production bottlenecks since

such bottlenecks have a MULTIPLICATIVE EFFECT on other production.
Suppose that n tasks are required to produce a good. Let q be the standard
skill level of these n tasks. But now let the actual skill level of two workers
be cut in half in all firms. With an O-ring production function, output would
fall by 75% But then the marginal product of quality also falls by 75% for
all the remaining n - 2 tasks, and thus so does the incentive to invest in
increasing skill.
A nation can get caught in economy-wide low-production-quality traps.
This will occur when there are O-ring effects across firms & within firms.
 Because there is an externality at work, there could thus be a case for
an industrial policy to encourage quality upgrading, as some East
Asian countries have undertaken in the past.
The O-ring result of Positive Assortative Matching relies on some rather
strong assumptions. The model’s basic results will follow so long as:
a) Workers are sufficiently imperfect substitutes for each other and
b) There is sufficient complementarity of tasks.
But how important are each of these assumptions?
4-28
Economic Development as Self-Discovery

Ricardo Hausmann & Dani Rodrik: A Problem of Information

It is not enough to say developing countries should produce “labor
intensive products,” because there are thousands of them.
Industrial policy may help to identify true direct and indirect domestic
costs of potential products to specialize in, by encouraging:
Exploration in first stage
Movement out of inefficient sectors and into more efficient sectors in
the second stage
THE GROWTH DIAGNOSTICS FRAMEWORK: focuses on a developing
country’s most binding constraints of economic development:
low rate of return on investment and high cost of financing
No “one size fits all” in development policy of market coordination
Insufficient investment in physical, social, and human capital
It is a valuable tool for domestic and international analysts who start
with a detailed understanding of a LDC; it can be helpful in identifying
binding constraints and the policy priorities to address them.
4-29
The Growth Diagnostics Framework

4 - 44
CONCLUSION
 The analysis of coordination failure problems in this chapter offers
some important overall lessons for policy, and indeed they present
something of a two-edged sword.
On one side, the analysis shows that the potential for market failure,
especially as it affects the prospects for economic development, is
BROADER and DEEPER than had been fully appreciated in the past.
E.g., the interactions of slightly distorted behaviors by potential investors
failing to consider the income effects of the wages they pay may produce
very large distortions, such as the outright failure to industrialize.
The coordination failures arising in the presence of complementarities
highlight the potential benefit of active government role in the context
of multiple equilibria to move the economy to a preferred equilibrium or
even to a higher growth rate that can then be self-sustaining.
The other edge of the sword is that with deep interventions, the
potential costs of a public role become much larger: policy choices are
more momentous because a bad policy today could push an economy
into a bad equilibrium for years to come.

4-31
CONCLUSION
This is because government can be a major part of the problem, playing
a key role in perpetuating a bad equilibrium such as a high-corruption
regime, in part because some government officials and politicians may
benefit personally from it.
Bad policy can even initiate a move to a worse equilibrium than a
country began with.
Both GOVERNMENT FAILURE and MARKET FAILURE (including coordination
problems & information externalities) are real, but public- and private-
sector contributions to development are also vital.
 Therefore, we need to work toward the development of institutions in
which actors in the public and private sectors have incentives to work
productively together (directly & indirectly) in such a way as to create
the conditions necessary to break out of poverty traps.
 In achieving this goal, the international community also has a vital
role to play, providing ideas & models and serving as a catalyst for
change, as well as providing some of the necessary funding.

4-32