Lewis' Dual Economy Model
Sources:
W.A. Lewis (1954) “Econonomic Development with Unlimited Supplies of Labour” Manchester School (see especially his summary points pp. 448-49)
Wikipedia on Dual Sector model: is quite good on the basics
D. Gollin (2014) “The Lewis Model:a 60 year Retrospective” Journal of
Economic Perspectives, v. 28, 71-88.
Paul Krugman “Hitting China’s Wall” NY Times July 18, 2013 (Chinese development and its 2013 situation viewed through Lewis’ model)
- Dualism:
- View that LDC economies have two quite different sectors:
- ‘Traditional’ and ‘Modern’
- Sectors are distinguished by:
- economic organization and institutions
e.g, importance of market exchange.
Traditional: subsistence sector;
Modern: capitalist, market-driven.
- state of technology
- role of capital:
- traditional: little capital lots of labour;
- modern: more capital per worker, returns to investment
relatively high.
- Usually discussed as if:
Traditional = agriculture = rural
Modern = industry = urban
- This need not be:
- modern agriculture possible (cash crop production)
- smallretail vendors, personal service (servants) in urban
areas: also traditional?
- Lewis model: economic development involves a transformation of the
economy from one dominated by the traditional sector to one
dominated by the modern sector.
- this is what is uniquely important about it.
- like a special case of the resource allocation model.
- mobilelabour between sectors;
- add a story about capital accumulation in the Modern sector;
- this gives the Lewis model.
- Lewis model and our earlier models?
- Traditional sector: is in a world much like the Malthusian model.
- population growth has kept income (wages) at subsistence levels.
- labour is in surplus: like the Harrod-Domar model
- Capital accumulation drives growth: like Harrod-Domar and Solow models.
Lewis model
- Created in the 1950s.
- A key model of development economics.
- Intended to fit the situation of a poor country with ‘overpopulation’.
- Focus on the transformation of the economy as it develops.
- Hybrid of the microeconomic model of resource allocation and a Classical/Malthusian
model.
- Classical aspects:
- labour supply is quite elastic: ‘unlimited’ at subsistence levels in the traditional sector.
- this keeps wages low and speeds the transformation.
- savings assumptions: capitalists save.
- capital accumulation is the key to growth.
(Inspiration? Poor countries in 19th century Europe )
- It is a model of the changing allocation of labour between traditional and
modern sectors.
Modern sector
- Two factors of production: capital and labour.
- Diminishing returns: given the amount of capital the value of marginal product curve for labour(VMP) will be downward sloping.
- Employers (capitalists)
- Assume that they are "wage takers" (competitive labour market)
(modern sector wage = WM)
- To maximize profits hire more labour as long as VMP>WM, i.e. hire labour up
to the point where:
WM = VMPM(LBest in the diagram)
- Represent the entire sector by an aggregate VMPM curve (summed across employers)
- Area under the VMP curve from 0 to LBest is the total value of output
produced.
- Total wages paid: WM LBest
- Area under the VMP curve minus WM LBest is the capitalists' (employers')
income.
- Capitalists are assumed to have very high savings rates.
- Most of their income goes into creation of more physical capital.
- Savings finances investment in the modern sector.
e.g., reinvestment of profits in own firm.
- Classical assumption, e.g., Marx: capitalists have a "drive to accumulate".
- As more capital is created in the sector the VMPM curve shifts up.
- Why?each worker has more capital to work with
labour is more productive
VMP rises.
Traditional Sector
- Say it is agriculture: Labour and land is used to produce output.
- Key assumptions?
- Productivity of labour in the sector is very low (initially).
- extreme diminishing returns: abundant labour, relatively little land.
(VMP very low)
- removing labour from the sector will have little effect on output in
the sector.
i.e., labour is surplus (like in Harrod-Domar!)
- very flat, very low VMPT curve when labour force is high.
- VMPT rises little or not at all initially as labour is withdrawn from the
sector.
- How did the sector get into such a state? Some kind of Malthusian
population growth story has resulted in ‘overpopulation’.
Labour market organization in the traditional sector:
- Two possibilities:
(1) Competitive market for agricultural labourers:
- Landowners hire agricultural workers.
- Competitive market (like modern sector):
WT =VMPT
- Labour income in the sector: WTL
- Remaining income goes to landlords.
- Landlords assumed to have low saving rates (say 0):
- spend their incomes on luxuries, servants, ceremony, etc.
(2) Alternative model of traditional sector labour market:
- Communal, (extended) family model.
- All family, clan etc. members can work in the sector.
- All are guaranteed some traditional level of subsistence.
- not necessarily paid VMPT
- sharing occurs to ensure at least subsistence.
- Extreme version: complete sharing
thenWT = VAPT
VAPT = Value average product of labour
i.e. Value of output/(No. of workers)
WT = worker-farmer’s traditional sector income.
- When labour is abundant:
VAP curve is flat at a low value (subsistence)
- Saving?
- All income goes to worker-farmers but VAP is at subsistence so
virtuallynosaving.
Transformation of the Economy
- Labour is assumed to be mobile between the two sectors.
- Labour migrates to the sector which gives the highest living standard.
- Usually assumed that there is a persistent wage gap between the two sectors.
- The modern sector pays a wage premium: D
- based upon empirical observation: Modern sector seems to pay more.
- this is may be an equilibrium or compensating differential: e.g.,
- compensation for living cost differences
- non-monetary costs of working in the modern sector.
- premium needed to cover mobility costs.
Lewis extra Modern sector pay compensates for:
“psychological cost of transferring from the easy going way of life of the subsistence
sector to the more regimented and urbanised environment of the capitalist sector.”
- Short-run equilibrium:
- Labour migrates between the two sectors until:
WT + D = WM
- The wage gap (WM-WT) is equal to the differential needed to attract workers
into modern sector.
(see for example pt. 0 in the following diagram when VMP0M is the VMP
curve in the Modern sector and VMPT is the VMP curve in the Traditional sector)
- Growth and movement to a new short-run equilibrium:
- Capitalists save and reinvest their income in the modern sector.
- Capital stock in modern sector now grows.
- So: VMPM curve shifts up to (VMP0M to VMP1M)
WM rises temporarily (to WMtemp given labour allocation is L0)
WT + D < WM
Migration from traditional to modern sector
WM falls as migration occurs (economy moving from pt. 1 toward
pt. 2)
Migration stops once: WT + D = WM(at pt. 2: new allocation of labour is at L1)
- At the end of this stage:
- morelabour in the modern sector
- wages are unchanged (note role of flat VMPT or VAPT)
- capitalists' incomes have risen.
- savings rate has risen since capitalists share of income is rising.
- More saving and reinvestment occurs the process repeats itself.
- VMPM for modern sector continues to shift up
- Modern sector grows, traditional sector shrinks
- But traditional sector output may not fall much: low productivity.
- Output of the two sectors combined is rising:
- difference of the VMPs on each unit of L transferred.
- Eventually the "turning point" is reached.
- Point where VMPT or VAPT starts to rise significantly.
i.e., surplus labourin the Traditional sector is exhausted.
- Further shifts in VMPM will cause a rise in wages in each sector.
- temporary boom in modern sector wages is not entirely reversed.
- rising productivity in traditional sector means the traditional sector
start to pay more: labour no longer in surplus!
- This will slow and perhaps stop the shift of labour.
Rising wages less income for capitalists less extra saving
- Also as capital grows in the modern sector more savings will go to
replacing worn out capital.
- Capital stock may eventually stabilize.
- It is possible that the traditional sector will be completely absorbed.
(e.g. if reach VMP5M in diagram)
Summary of the model's predictions
- Transformation of the economy will occur.
- traditional/agriculture shrinks.- modern industry grows.
- Transformation is driven by growing incomes of capitalists and capital
accumulation in the modern sector.
- Savings rates will rise during the development process.
(notice reliance on capitalists getting much of the extra income)
- Wages will not rise during the transformation:
reflects surplus traditional sector labour
- early Industrial Revolution an example.
- Income inequality will rise during the development process.
- After the turning point, rising wages will slow and possibly stop the
transformation.
- End result: transformation slows and ends.
- Growth from reallocation ends.
- No longer a surplus labour economy. Rising wages.
- Krugman “Hitting China’s Wall”
- Is China at or approaching this point?
- Fast growth since early-1980s partly reflects transformation.
- moving resources from low-productivity sectors to higher
productivity sectors.
Complications and Criticisms
- Transformation from traditional to modern is key to development.
- Is it automatic?
- Can government or aid spur this transformation?
- Some have transformed some have not: why?
- Acemoglu and Johnson Why Nations Fail Ch.9
- Why a dual economy?
(discussion of South Africa as a created dual economy;
plantation economies; Latin America as well?)
- Does it reflect political and economic institutions?
- Is transformation impossible without different institutions?
- Model: Capitalists save their income domestically.
- demonstration effects: will they instead use their incomes to import
luxuries from MDCs?
- capital flight: will they send savings abroad?
where are the highest and safest returns?
- Terms of trade between sectors:
- Model above does not discuss what happens to prices of traditional and
modern sector output.
- These prices also determine: VMPM and VMPT
- A sharp rise in price of traditional output vs. modern sector output could
slow or halt the transformation.
- low productivity in agriculture: early in the process supply of
traditional output changes little so price may be stable.
- modern sector output:
- could price fall as sector expands?
- less likely if the output is exported.
(does the model best fit countries who are exporting modern sector output? Otherwise who buys the output?)
- Modern "enclave" as an equilibrium outcome
- A case of stalled transformation.
- Why might it stall?
- Terms of trade between sectors.
- Modern sector may be resource based:
- limited by resource availability.
- Wages early-on rise for reasons outside the model: policy, unions?
- Traditional sector is passive in this model.
- This sector plays a passive role: provides a supply of labour to the modern
sector.
- extralabour helps create capitalist's income, generates savings, this
drives development.
- Policies:Anti-traditional sector bias
e.g., encourage industry (modern)
neglect agriculture (traditional)
- taxation or marketing boards can be used to extract surplus from
traditional sector.
- too much extraction: famine, weak incentives to produce.
(Soviet Union, China, African examples in Acemoglu & Johnson)
- Traditional sector: viewed as custom-tradition bound, decision-making little
affected by incentives.
- evidence suggests the contrary.
- Idea that saving and accumulation can from the rural sector.
- could also fuel investment.
- A healthy agricultural sector can be important to development
- urban wages rise with the price of foodstuffs
- China since 1978: agricultural reforms important early in the process.
- village levelenterprises.
J. Studwell(2013) How Asia Works : on East Asian success stories claims
that agricultural reform was important in all successful cases.
- Surplus labour assumption
- Early version of the model claimed VMP=0 in the traditional sector:
this appears to be wrong.
- how flat is the VMP curve in the traditional sector
- wages often not as stable as the model predicts.
- Empirical support for agriculture having low productivity vs. industry receives
much support.
1