Is India’s manufacturing sector moving out of the cities?

Ejaz Ghani, Arti Grover Goswami, William Kerr, 27 July 2012

a

A

By 2025, India will have more than double the number of city dwellers than the United States has today (McKinsey 2010). Urbanisation is central to the development process. Urbanisation, industrialisation, and economic development are expected to go hand in hand as employment shifts out of agriculture, and industrial production is located in cities with their agglomeration economies (World Bank 2009). However, while urbanisation has moved ahead at a rapid pace, the pace of industrialisation has slowed down (Henderson 2012). The disconnect between urbanisation and industrialisation raises a number of questions. Is the link between urbanisation and industrialisation more complex than what we had earlier thought? Why are formal firms moving out of cities? And why are informal firms moving into cities? Do the agglomeration benefits of density and city differ across formal and informal sectors?

In a recent paper (Ghani et al. 2012), we investigate the urbanisation of Indian manufacturing by combining enterprise data from formal and informal sectors. Our study suggests that plants in the formal sector are moving away from urban into rural locations, while the informal sector is moving from rural to urban locations. Although the secular trend for India’s manufacturing urbanisation has slowed down, the localised importance of education and infrastructure have not. Districts with better education and infrastructure have experienced a faster pace of urbanisation, although higher urban-rural cost ratios cause movement out of urban areas. This process is associated with improvements in the spatial allocation of plants across urban and rural locations.

Figure 1a. India's urban shares, 1989-2005

Notes: Figure plots urban shares of plants, employment and output for each year using survey data of plants from organised and unorganised sectors.

Figure 1a shows that India’s manufacturing sector has become more urbanised, with the share of workers in urban areas rising from 33% of employees in 1989 to 41% in 2005. This pattern of increasing urbanisation is also observed when looking at manufacturing plant counts, although the trend is reversed with manufacturing output. The latter has increasingly moved towards rural areas.

Figure 1b. Urban shares in organised sector

Figure 1c. Urban shares in unorganised sector

The differences in the relative movements of formal and informal firms are striking (see Figures 1b and 1c). The organised sector moved from urban to rural locations, with its urban employment share declining from 69% in 1989 to 57% in 2005. Per contra, urban employment share for the unorganised sector increased from 25% to 37%. Since the unorganised sector accounts for about 80% of employment in India’s manufacturing sector, the total urbanisation level increased for the employment measure. Likewise, the organised sector accounts for over 80% of India’s output, such that the aggregate output series instead became more rural.

Organised and unorganised sectors

The distinction between organised and unorganised sector in the Indian context relates to establishment size. In manufacturing, the organised sector is comprised of establishments with more than ten workers if the establishment uses electricity. Establishments are surveyed in India with state and four-digit National Industry Classification (NIC) stratification. These surveys identify for each establishment whether or not it is in an urban or rural location.

India uses a more demanding set of criteria than most countries to define what is ‘urban’. For instance, substantial parts of US metropolitan areas such as Atlanta or Phoenix would be classified as rural in Indian statistical analyses because their population densities fall below 1,000 persons per square mile. Thus, our measured urbanisation for India will be lower than many international standards. This consideration does not affect, however, the longitudinal consistency of our trends for India.

Table 1. Estimations for 1994-2005 urbanisation changes in manufacturing sector

Analysis of urban movements

To assess the factors contributing to changes in India’s urban shifts, we consider a series of regressions that quantify district traits that are associated with increased urbanisation. Table 1 investigates the correlation of India’s urbanisation of employment at the district-industry level from 1994 to 2005. We find substantial evidence that links greater urbanisation to districts with more educated workforces and better infrastructure levels. We also find evidence that higher costs, or sharper differences in urban-rural cost levels, decrease the rate of urbanisation (columns 3 and 4 in Table 1). These effects are most pronounced in the unorganised sector and before 2000.1

Anecdotal accounts of India suggest multiple relationships are at play. For example, a natural baseline for education is that agglomeration economies or urbanisation premiums are higher for skilled workers and industries. Observers note, however, that skilled workers may want to live outside of cities to the extent that the amenities are lower in Indian cities than in the surrounding areas. Likewise, better infrastructure typically allows strong urbanisation levels. But, cities in India often experience some of largest infrastructure failures, and production of own electricity for organised sector firms is high in both settings. Thus better infrastructure capacity may allow establishments to move to rural locations.

Observers have frequently noted the relatively slow pace of India’s urbanisation (even recognising the differences in urban definitions); moreover, the movement of organised manufacturing sector plants to rural areas is surprising, given the relative youth of India’s manufacturing sector. Perceived wisdom is that this sluggishness is in part due to the limits imposed by India’s poor infrastructure and weaker education levels, among other factors like strict building regulations. Our work supports these claims. Continued investment in these factors, beyond their direct effects for Indian businesses, will also provide beneficial effects from an urbanisation and spatial allocation perspective.

Finally, our analysis indicate that higher build-up, especially in the top quartile, is strongly associated with increased urbanisation (columns 7 and 8); we find similar effects with other specification formats like linear shares. A plausible interpretation of these coefficients is that land availability is a strong governor of the urbanisation process. While this interpretation would match anecdotal accounts of the constraints that land availability and real estate prices have for the location choices of manufacturing firms in India, this measure is indirect (e.g., compared to real estate price data) and thus caution in interpretation is warranted.

One effect of adding this control is that the coefficient on the infrastructure index declines somewhat and becomes insignificant when including state fixed effects. This is not too surprising given that the built-up metric in part captures higher infrastructure levels. Adding this variable, however, does not affect the results on our wage variables. We take from this set of results that costs factors that slow manufacturing urbanisation in India are definitely present in the labour market and most likely present in the real estate market, too. The evidence further suggests that absolute cost levels and localised urban-rural cost ratios play a role in India’s urbanisation.

The combined effect of multiple layers of poorly conceived central, state, and municipal regulations contribute to an artificial urban land shortage in India. As a result urban land prices are abnormally high in relation to India’s household income, and households consume less floor space than they could afford if the regulatory environment were reformed. In addition, some regulations have a negative impact on the spatial structure of cities. By unreasonably reducing the amount of floor space that can be built in centrally located areas, and by making land recycling difficult, some regulations tend to “push” urban development toward the periphery. As a result, commuting trips become longer, public transport become difficult to operate and urban infrastructure has to be extended further than what would have been the case if land supply had been unconstrained.

Conclusions

Spatial location of plants has implications for policy on investments in education, infrastructure, urbanisation, and the livability of cities. It is very important for policy makers to recognise that much of the urbanisation that is occurring in India is in the unorganised sector. Policy makers should take an inclusionary, rather than exclusionary, approach to the urban informal economy.

Going forward, there should be adequate provision of infrastructure for the informal sectors to develop. Indian cities should find ways to ensure that urban informal livelihood's are integrated into urban plans, land allocation, and zoning regulations; that the urban informal workforce gain access to markets and to basic urban infrastructure services; and that organisations of informal workers are invited to participate in government procurement schemes and policy-making processes. The more that the Indian cities recognise this influx and design appropriate policies and investments to support it, the more effective the policy interventions will be.

References

Ghani, Ejaz Arti Grover Goswami, and William R Kerr (2012), “Is India's Manufacturing Sector Moving Away From Cities?”, NBER Working Paper No. 17992.

Henderson, J Vernon (2012), “Cities and Development”, Journal of Regional Science, 50:515-540.

McKinsey Global Institute (2010), “India’s urban awakening: Building inclusive cities, sustaining economic growth”, April.

Planning Commission (2008), “Eleventh Five Year Plan 2007-2012”, Planning Commission, Government of India, New Delhi.

World Bank (2009), Reshaping Economic Geography, World Development Report.


1 Results in columns 5 and 6 assure us that our results are not driven by reclassification of rural areas (by controlling for district land area and change in urban population to the preceding specification

Topics: Development, Industrial organisation
Tags: India, industrialisation, urbanisation

Ejaz Ghani
Lead Economist in Economic Policy and Debt, PREM Network, World Bank
Arti Grover
Consultant, International Trade Department, World Bank
Assistant Professor, Harvard Business School