Lecture2: Migration and Urban Unemployment : The Harris-Todaro Model and Policy Issues

• Urban unemployment is an issue, which Lewis model does not consider. Further, the conventional
remedies for urban unemployment have often surprised governments by worsening rather than Michael P. Todaro

ameliorating the unemployment problem. One instance is reported in Todaro (1969). It has been (1969). “A Model

observed in times that attempts to remove urban unemployment directly by creating more jobs, has had of Labor Migration
and Urban
the opposite effect.
Unemployment in
• According to Harris and Todaro, creating urban jobs by expanding industrial output is insufficient
Less Developed
for solving the urban unemployment problem.
Countries”. In:
• For these reasons, it has been argued that, it is important to construct a realistic theory of rural-urban The American
labour distribution that can be used for policy purposes. And such a theory needs to be based on some Economic Review

hypothesis of migration. 59(1), pp. 138–148.

• Instead they recommend that government reduce urban wages, eliminate other factor price
distortions, promote rural employment, and generate labour-intensive technologies, policies
• The underlying motivation behind migration in Haris-Todaro model is the economic
motive of improving one’s material living conditions.
• Suppose there are L workers in the economy, with LM and LR being the numbers
employed in the modern and rural sectors. The urban wage is fixed at w¯ , and the
rural marginal product of labour (which is assumed to be equal to rural wage) is
fixed at wR, where wR < w¯.
• Assume that the number of urban jobs available (LM ) is exogenously fixed. This
implies that if there are more workers than LM in the urban sector, they would have to
be unemployed, whereas in the rural sector labourers can always find work.
• The total urban labour force equal L − LR with (L − LR ) − LM being unemployed.
 HT Model: Simple version
• The crucial assumption of the HT model is that workers base their migration decision on their
expected incomes. Note that finding a job in rural sector is assured, and hence expected rural
income is wA .
• Expected urban income is determined by multiplying w¯ by the probability of finding a job
there, which the model assume to be equal to the rate of urban employment, LM /(L− LR ).

• Hence, as long as 𝑾 𝑳𝑴 > 𝑾𝑹 #will be rural-urban migration.

𝑳−𝑳 𝑹

ഥ and 𝑊𝑅 are fixed, as people migrate from villages to the modern sector, that
• Since, L, LM, 𝑊
is as LR falls, the left-hand expression falls as well.
𝑳𝑴
• Equilibrium is reached when there is no tendency to migrate, that is when 𝑾
𝑳−𝑳𝑹
= 𝑾𝑹
 HT Model:Simple version

The 1964 Kenyan experience mentioned above is capture well even in this simple model.
Rewrite the equilibrium condition as follows:

w¯
LR = L − LM
wR
hence
∂LR w¯
=−
M Parameswaran

∂LM wR

This implies that if the number of urban jobs is raised by one unit, rural employment falls by
w/wR units. In other words, creating one additional job in the urban sector induces w¯/wR
people to migrate to the urban sector. Since w̄ > wR , this means urban unemployment instead
of, declining, rises.
• In this model, creating urban employment invariably increases
urban unemployment and the complete HT model is flexible in
this respect.
• Also, LM and wR are endogenously determined in the actual HT
model, though w¯ continuous to be exogenously determined.
So we now present the complete HT model.
 HT Model: Complete Version

• There are two sectors, rural (R) and urban or modern (M). They produce XR and XM units of output and employ

LR and LM units of labour force.

•The model is a short run model, so capital stock is fixed. Hence output is a function of the labour input. And

the production functions are respectively

• XR = fR (LR ) 𝑓𝑅′ > 0; 𝑓𝑅′ < 0

𝑗 𝑗
• XM = fM (LM ) 𝑓𝑀 > 0; ; 𝑓𝑀 < 0

• Total labour units available in the economy is fixed at L. Therefore,

• LR + LM ≤ L LR ; LM ≥ 0
• Let us assume that both sectors produce same good, though by different techniques. This
assumption helps us to avoid the price formation equation, and major results of the model
remain unaltered even if we assume both sectors produce different goods. OR the units of XM
and XR can be expressed in such a way that their prices are equal to one.
• Let w be the urban wage in real terms, and urban capitalists is an wage taker, and objective is
to maximise profit. This implies following
𝑗
𝑓𝑀(LM ) = w

• Urban wage has political or institutionally determined lower bound w¯, so

w ≥ w¯
•This is an important assumption and we denoted it as “wage rigidity axiom”.
• It is assumed that w¯ is above the wage rate that would have prevailed if the
minimum wage rate is not fixed. This ensures that w¯ is above the equilibrium
wage rate and there is excess supply of labour.
• Hence profit maximising condition implies that
𝑗
𝑓𝑀 (LM ) = w¯

• In the rural sector, wage rate is flexible and equal to the rural marginal product fR𝘫
(LR ). This flexibility ensures that there is no rural unemployment. So the total
work force, both employed and unemployed, in the urban sector is equal to L − LR .
 HT Model: Complete version

The probability that a migrant labour find an urban employment is

LM
L − LR

Workers maximise expected earnings, hence if

LM
𝒇′𝑹 (LR ) < w̄
L− LR workers would migrate from rural to urban
M Parameswaran

sector.

And migration equilibrium is attained when

L
𝒇′𝑹 (LR ) = w̄ − ML
L R

This completes the HT economy.

Note that the above condition alone cannot ensure an equilibrium in the HT economy. An HT
equilibrium is a situation satisfying following five conditions.
𝑗 𝑗
XR = fR (LR) 𝑓𝑅 > 0; 𝑓𝑅 < 0

𝑗 𝑗
XM = fM (LM ) 𝑓𝑀 > 0;; 𝑓𝑀 < 0

M Parameswaran 𝐿𝑅 + 𝐿𝑀 ≤ L LR;LM ≥ 0
𝑗 ഥ
𝑓𝑀 𝐿𝑀 =𝑊
𝑳
𝒇′𝑹 𝑳𝑹 =𝒘
ഥ 𝑴
𝑳−𝑳 𝑹

Given the above five equation it is possible to solve for LM , LR , XM , and; XR . Let us denote this
solution as
0 , L0 , X 0 , X 0 ], let us denote this by E 0.
[LM R M R

Since E0 satisfies all the five conditions above, it is an HT equilibrium.

• Before further discussing this equilibrium, let us characterise this equilibrium. Assume that
social welfare, U depends on the output produced, namely XM and XR . Since both sectors
produce same goods, we write

U = XM + XR

•Assume a command economy, in which the government can distributed the labour force between
the two sectors, R and M as it wishes and the only constraint it faces is given in the above first
three equations.
• In this set up the government chooses LR and LM by maximising the utility function subject to
the first three constraints. Since marginal products in both sectors are greater than zero,
unemployment is never desirable.
• Hence, the 3rd constraint holds as strict equality.
• In these conditions, maximisation is a Lagrangian exercise that yields the following first-order
conditions.

LR + LM = L

𝑗 𝑗
𝑓𝑀(LM ) = 𝑓𝑅 (LR )

• Let 𝐿∗𝑀, and 𝐿∗𝑅 be the solutions to the above equations. By inserting these values in the
∗
production function we get the optimum sectoral outputs 𝑋𝑀 , and 𝑋𝑅∗.
• Now we can see that if there is no wage rigidity axiom, the optimum level of outputs coincide
with free market outcome or laissez-faire outcome.
ഥ the HT equilibrium is suboptimal:
• However, due to wage rigidity (𝑊),
Urban unemployment exists.
Marginal productivity of labor in the urban sector exceeds that in the rural sector.
Harris-Todaro Migration Model
• Since H is a rectangular hyperbola,
ഥ 𝑂𝐿𝑂𝑀 =(OQ ) (O𝐿𝑂𝑅 )
𝑤
• This implies that 𝑤 ഥ 𝑂𝐿𝑂𝑀 /O𝐿𝑂𝑅 =OQ

• This is the equilibrium condition given above. Hence 𝐿𝑂𝑀

and 𝐿𝑂𝑅 in figure represent HT equilibrium. Urban
unemployment consist of 𝐿𝑂𝑀 𝐿𝑂𝑅

• What is theoretically a vexing issue is the formation of

urban wage. In HT model and in all early dual economy
models, urban wage are simply assumed to be rigid
downward.

• This ensures that in equilibrium ഥ >𝑓𝑅′ 𝐿𝑂𝑅

𝑊

• This in turn implies urban unemployment, otherwise the

probability of getting an urban job is equal to 1. Hence, the
expected urban earning is equal to w̄ and this exceeds the rural
wage 𝑓𝑅′ 𝐿𝑂𝑅 ; But this is incompatible with equilibrium
Options open to a potential migrant

• The left set is a single box: agriculture, with its wage 𝒘𝑨

• The right set : the probabilities of access in formal sector at some
high wage: 𝑤
ഥ.
• The probability of obtaining formal job = p.
• wage rate in the informal sector by 𝑤𝐼
• probability of getting an informal sector job, conditional on
having been turned away from the formal sector: q
• remains openly unemployed with probability 1 − q
Therefore: q 𝑤𝐼 + (1 − q)0 = q 𝑤𝐼
Thus the overall expected wage = p + (1 − p) q 𝑤𝐼

LF + LI = L
 Policy Intervention

• H.T recommends Dual Policy requirement

• Bhagwati and Srinivasan (1974) proposed a dual policy:
1. Subsidise urban wages to achieve optimal urban employment (to employer, worker get high w¯ only).
2. Simultaneously subsidise rural employment to shift the marginal product curve upward, achieving optimal rural
employment. (Restrict rural Migration)

•The equilibrium in the HT economy is sub-optimal, not only is it characterised by unemployment, but
the marginal labour in the urban sector is more productive than the marginal labour in the rural sector.
• Now the question what policy measures can be taken to shift the economy toward the optimum 𝐸 ∗ ?
• First, consider an employment subsidy of s to the urban sector, so the new urban wage become w¯ − s.
This means that the urban unemployment is equal to OK (in figure).
• For the same reason mentioned earlier, urban unemployment may increase.
 Figure: Case of employment subsidy

M Parameswaran
• From the figure, it is easy to see that E ∗= [L∗M, L∗R, XM∗ , XR∗] is unattainable by the urban
employment subsidy.
• To induce urban firms to employ the optimum amount L∗M , a subsidy of s∗ defined
below is needed (see the figure)
s∗ ≡ w¯ − fM (L∗M )

• Given this urban wage-employment point depicted by N, the equilibrium rural

employment is OJ , which is less than the optimal O 𝘫L∗R.
• The above point implies that an urban subsidy that ensures optimal employment in the
urban sector, necessarily implies sub-optimal employment in the rural sector.
• Therefore, it has been argued that no single policy could correct the sub-optimality of
the HT economy.
• One option is to have combinations of different policies, like S∗ subsidy to
the urban sector and simultaneously restricting migration from rural to Jagdish N.
Bhagwati and T. N.
urban, so that only workers who got a job in the urban area is allowed to Srinivasan (1974).
“On Reanalyzing
migrate. However, this policy is ethically unappealing and politically not the Harris-Todaro
Model: Policy
feasible.
Rankings in the Case
• Another option is to give subsidy to rural workers also. If subsidy is of Sector-Specific
Sticky Wages”. In:
given to the rural workers MPR would be shifted upwards and The American
Economic Review
employment in the rural sector increases. 64(3), pp. 502–508.

• A subsidy of S∗ in the rural sector ensures that optimum level of

employment is reached. That is give both rural and urban workers a
subsidy of S∗. This is the central result of Bhagwati and Srinivasan (1974).
• The major issue here is information. How to compute the subsidy amount, which is
equal to S∗ = w¯ − fM 𝘫 (L∗M ).
• To be able to compute S∗ we need to know f 𝘫(lM∗ ), that is the marginal product of labour
that would prevail at the optimum. At the time of implementing the policy, this is
unknown.
• However, it can be shown that a uniform subsidy, equal to greater than S∗ is optimum, that
is even a subsidy of S > S∗ is given, a labour distribution [L∗M , L∗R ] is attained.
 Real-World Cases: Harris-Todaro Model in Action

1. India: Post-COVID Migration

5. Latin America
•Urban job losses → Rural-to-urban migration
•Brazil, Mexico: Urban informal jobs dominate.
resurgence.
•Insight: Dual economies persist with migration pressure.
•Insight: Urban wage expectations > Rural incomes.
6. Southeast Asia
2. Sub-Saharan Africa
•Jakarta, Manila: Urban slums and underemployment.
•Youth unemployment > 30% (Ghana, Kenya).
•Insight: Urban wage premiums attract migration despite risks.
•Insight: Rural push factors sustain urban migration.
3. Middle East: Refugee Dynamics
Policy Matters:
•Urban informal jobs grow (Jordan, Lebanon).
•Strengthen rural development to reduce migration pressure
•Insight: Migration outpaces formal job creation.
(India’s MGNREGA).
4. Ethiopia: Industrial Parks
•Recognize and formalize urban informal sectors.
•Hawassa: Migrants exceed formal job creation.
•Promote regional balance and infrastructure investment.
•Insight: Wage gaps drive migration; informal sector
•Align urban wages with productivity to reduce unemployment
absorbs surplus.
gaps.