Co-location of Formal and Informal Manufacturing and Firm’s Performance: Evidence from India

Abstract

This study analyses the impact of co-location between formal and informal manufacturing sectors on plant-level productivity. We employ a unique data obtained by merging plant-level data from Annual Survey of Industries (ASI 2011–2012) and Survey of Unorganised Manufacturing and Repairing Enterprises provided by National Sample Survey Office (NSSO 67th round). We find that formal and informal manufacturing plants gain from localisation. Further, co-location with informal enterprises has a positive impact on productivity of formal sector plants; however, we observe insignificant impact of co-location on informal sector enterprises. Additionally, we find evidence that informal sector enterprises benefit from industrially diversified regions.

JEL Codes: D24, R12, R3

Keywords

Productivity agglomeration informal sector

1. Introduction

Spatial concentration of the manufacturing sector and the associated productivity benefits for firms are well established in the theories of agglomeration. Marshall (1920) attributes externalities within the industry, including access of specialised inputs, availability of labour pool and knowledge sharing, as the key drivers of productivity of regionally concentrated firms. Later, Jacobs (1969) explains industrial diversification of the region as another important determinant of a firm’s productivity. According to Jacobs (1969), by the virtue of being located in an urbanised region, firms gain from various externalities associated with the cities such as better technological and transportation facilities, specialised financial and professional services, inter-industry information transfers and highly developed markets. In other words, it refers to the externalities within the region, irrespective of the industries. Based on these arguments, multiple studies have analysed the impact of industrial location on a firm’s productivity, more specifically, the impact of agglomeration economies on the firm’s productivity (Ciccone & Hall, 1996; Lall et al., 2004; Maré & Timmins, 2006; Overman et al., 2010).

However, most of the existing literature on agglomeration economies and productivity is confined to the experience of the formal manufacturing sector and largely neglects the informal sector. Since rapid urbanisation in developing countries integrates a large number of informal firms and informal workers in cities (Ghani & Kanbur, 2013), it is imperative to analyse the spatial distribution of the informal sector and its impact. Informal sector accounts for a large share of manufacturing employment in almost all developing economies, including India (Jutting & de Laiglesia, 2009). In the case of India, more than 80% of manufacturing workers are engaged in the informal sector (ILO, 2011–2012). However, due to the lack of availability of high-quality data on informal sector enterprises, very few studies have attempted to analyse the regional distribution of the informal sector. In the context of India, existing studies have attempted to provide micro-level evidence on the spatial aspects of the informal sector (Ghani & Kanbur, 2013; Mukim, 2015), while another strand of studies have analysed the migration of informal sector from rural to urban areas (Ghani & Kanbur, 2013; Mukim, 2015). The results of these studies show geographical proximity of formal and informal sectors through labour–market, buyer–seller and technological linkages.

However, with the exception of Ghani and Kanbur (2013) and Mukim (2015), studies pertaining to the spatial distribution of the informal sector is relatively scant in the context of India. Therefore, in this article, we address two questions regarding the spatial distribution of the informal sector. First, whether informal sector benefits from regional concentration? Second, how formal and informal sectors benefit from co-location? While there is an abundance of studies evaluating the impact of agglomeration economies on the formal sector, those related to the agglomeration economies on informal sector are few in number. Even though Duranton (2008) states that agglomeration economies accrue to both formal and informal sector, the empirical evidence is still lacking. Tran and La (2018) report that the externality effects on the labour productivity is different in rural and urban regions. On the other hand, Overman and Venables (2010) find that high regional concentration of informal sector might push up the cost in urban regions, which may reduce the agglomeration effect. Therefore, the existing empirical evidence finds mixed effect of agglomeration on the informal sector. Similarly, there is no theoretical consensus about the effects of co-location since it can have positive and negative impacts. In the case of formal and informal co-location, Arimah (2001) asserts that co-location of formal and informal sector enterprises will lead to productivity improvements. Though Mukim (2015) analyses the co-location pattern of formal and informal sector, the productivity effects on formal and informal sectors are overlooked. A study closer to the current work was from Tran and La (2018) who report that co-location with formal firms generate negative impacts on informal productivity. Since, different studies posit differential effects of co-location, we believe that empirical verification of the impact of co-location on firm- as well as industry-level productivity, separately for formal and informal sectors, assume greater significance.

Based on the above-mentioned discussion, this article makes the following contributions to the literature. First, we identify the spatial concentration of the informal sector enterprises and analyse the impact of this concentration on the plant-level productivity. Since there is a large migration of informal sector to urban regions, it becomes crucial for urban and regional policymakers to understand the spatial pattern of the informal sector. Second, even though existing studies on the informal sector show unambiguously its linkage with the formal sector (Mukim, 2015; Ranis & Stewart, 1999; Venables, 1996), little is known about the co-location behaviour. Distinct from these studies, we bring new evidence on the impact of co-location of formal and informal sectors on plant-level productivity of both sectors.

The remainder of this article is organised as follows: the second section briefly reports the changes in the location pattern of formal and informal sectors in India. The third section introduces the data sources and explains the measures of industrial agglomeration. The fourth section explains the estimation procedures. The fifth section discusses the empirical results, and the sixth section concludes the article.

II. Changes in the Location Pattern of Formal and Informal Sector

In order to understand the location pattern of formal and informal sectors, we report state-wise changes in the formal and informal manufacturing employment share. In India, according to the Factories Act of 1948 and the Beedi and Cigar Workers Act of 1966, formal manufacturing includes those units or plants registered as factories under Sections 2(m)(i) and 2(m)(ii). Accordingly, a plant which employs more than 10 workers with power and more than 20 workers without power belong to the formal manufacturing sector. On the other hand, the informal sector includes all manufacturing enterprises, which are not registered under the above-mentioned Act. Further, there is no uniform definition of informal employment. The 15th International Conference on Labour Statistics (ICLS) defined it as all the persons who are employed in at least one informal unit (ILO, 1990). Consequently, 17th ICLS defined it as the employment, which is not covered under national labour legislation, income taxation, social protection or entitlement to certain employment benefits (ILO, 2003). Following the 17th ICLS conceptual framework on informal employment, National Commission on Employment in the Unorganised Sector (NCEUS, 2007, defined informal employment as ‘of those working in unorganised enterprises or households, excluding regular workers with social security benefits, and workers in formal sector without any employment/social security benefits provided by the employers’. Therefore, given the multiplicity of the definition of the informal sector, informal workers in India can be considered as all the workers engaged in any sector without any social security benefit (Abraham, 2017). According to the International Labour Organization (ILO, 2011–2012), around 92% of the manufacturing workers are engaged in the informal sector in India. This includes all the workers in the informal sector and the informal sector workers in the formal sector on contract. Similarly, Ghani et al. (2012) report an increase in the share of informal manufacturing employment in the urban regions. In a similar vein, in Figures 1 and 2, we provide changes in the state-wise share of employment among formal and informal sectors during the period from 2006 to 2011. It is evident from the figures that the state-wise distribution of employment share in the formal sector does not show much change. On the other hand, in the case of the informal manufacturing sector, we find a substantial change in the state-wise employment share. States like Maharashtra, Karnataka and Andhra Pradesh show substantial increase in the informal employment share along with Tamil Nadu, Tripura, Haryana and Uttarakhand. According to NSSO database, the increase in the informal employment in India is mainly attributed by the increase in the informal employment in urban areas. During the 2006–2011 period, the number of workers in the informal sector in urban areas recorded 26% increase. According to Ghani and Kanbur (2013), though there are agglomeration benefits for both formal and informal sectors, these benefits appear to be stronger for the informal sector in urban areas. Further, they argue that the co-movement of urbanisation and informality is not a surprise in India. In support of this argument, we also observe an increase in informal employment in developed states than the developing states. Similarly, during the period from 2001 to 2011, employment in the formal sector increased from 7.7 to 12.9 million. Interestingly, more than half of this increase is accounted by the upsurge in the contractual workers than the directly employed workers. During the same period, the share of contract workers in the formal sector increased from 15.7% to 26.5%. Therefore, the upsurge in the informal workers in developed states can be explained by the increased migration of informal workers to urban regions and the increasing share of contract workers. On the other hand, states like Jharkhand, Bihar, Rajasthan, and Jammu and Kashmir witnessed a decline. Further, not much changed was observed in certain states such as Kerala, Gujarat, West Bengal, Uttar Pradesh, Assam, Madhya Pradesh, Chhattisgarh and Odisha. Traditionally, a substantial share of formal employment is concentrated in Tamil Nadu, Maharashtra and Andhra Pradesh. Interestingly, the same states witnessed an increase in the informal sector employment (Figure 2), which indicates a sign of co-location of formal and informal sectors.

Figure 1.

Source: Authors’ calculation based on ASI database.

Figure 2.

Source: Authors’ calculation based NSSO survey.

III. Data and Variable Description

For this study, we have obtained information from two data sources. For the formal sector, we relied on the Annual Survey of Industries (ASI), and, for the informal sector, we used Survey of Unorganised Manufacturing and Repairing Enterprises provided by the National Sample Survey Office(NSSO). Since the NSSO enterprise survey data are quinquennial in nature, we use the 62nd round (2005–2006) and 67th round (2011–2012).¹ In order to achieve consistency in terms of the time period, among different annual surveys on the formal sector, we use ASI 2005–2006 and ASI 2011–2012.² ASI survey during the 2005–2006 period includes 57,304 plants and, similarly, 52,243 plants during the 2010–2011 period. NSSO provides information related to 80,637 and 141,744 enterprises, respectively.

Before proceeding to the empirical analysis, we have resorted to data filtering. Initially, our data set contains information pertaining to 109,548 plants, belonging to the formal sector and 206,100 plants in the informal sector. In the second stage, we consider only those entities that are operational during the survey period; therefore, we are left with 88,162 and 188,054 enterprises in the formal and informal sectors. The key variable of interest is ‘labour’ since the measurement of both productivity and agglomeration measures are dependent on this information. Therefore, we eliminate all the firms, which reported missing or zero values of labour, and as well as those firms which reported negative output and gross value added (GVA). Further to bring congruity with the National Industrial Classification (NIC) codes,³ we made a concordance between NIC 2004 and NIC 2008, and we used NIC 2008 classification for the empirical analysis. Finally, as mentioned earlier, an important challenge encountered in the data set is the unavailability of the district information. Given the fact that size of some of the states in India is similar to the smaller countries, a district-level analysis is more meaningful (Mukim, 2015). Even though NSSO surveys include district-level information of each enterprise, ASI provides the same information only up to the year 2008. From 2008 onwards, ASI provided only unique plant identifier, but it stopped providing the district identifier. Therefore, we follow Martin et al. (2017) to identify the district codes by using opening and closing values of different variables. We use the opening values of six variables in 2008 and match it with the opening values of 2009 and continue this process to the next year. These variables include stock of raw materials, fuels and stores, stock of semi-finished goods, stock of finished goods, inventory, loans and fixed capital. Further, we restricted our study period up to the year 2010 since further matching led to a huge decline in the number of observations, and we were unable to include the information regarding the new entrants. This may lead to an incorrect measurement of agglomeration variables. After the matching, we ended up with 26,157 (50.06% of total plants) plants in formal sector and 62,801 plants (44.30% of total enterprises) in the informal sector during the 2010–2011 period.

Variables

To analyse the effects of agglomeration economies on the plant productivity, we classify the agglomeration variables into two categories based on the definitions given by MAR externalities and Chinitz–Jacobs urbanisation, namely MAR localisation and Jacobs’ urbanisation (definition of variables are given in Table 1). Further, we measure localisation and urbanisation variables for both the formal and informal sector industries. MAR localisation indicates the clustering with other enterprises in the same industry. By clustering within the same industry, a firm can take advantage of information spillovers, access the specialised sector-specific inputs and skilled labour. Therefore, cost-saving externalities may maximise, and all the firms in the cluster may experience improvements in productivity. Empirical studies related to the localisation and productivity confirm a positive localisation externalities in both developed (Black & Henderson, 2003; Ciccone & Hall, 1996; Duranton & Puga, 2001) and developing economies (Lall et al., 2004). On the other hand, Overman and Venables (2010) argue that excessive clustering of firms in a region may lead to the problems of congestion and increased price for land and other inputs, leading to negative externalities.

There are different methods adopted to measure localisation.⁴ Henderson (2003) suggests that the count of own-industry enterprise is the best method rather than counting own-industry employment. Lall et al. (2004) employed location quotient index to measure the location in the Indian context. Following these studies, we employ the location quotient index to measure the localisation, measured as the rate of concentration of an industry in a region as compared to the total economy. We define location quotient as:

L_{j r} = \frac{E_{j r}}{E_{j}} / \frac{E_{r}}{E_{n}}

(1)

where E_jr is the total employment in industry j in region r, E_j is the total employment in industry j in the entire economy and E_n is the total employment in the economy. Since we include both formal and informal sector industries in our analysis, we classify the localisation index as follows: (a) measure the localisation of formal sector and (b) localisation of the informal sector. This classification helps us to analyse the importance of localisation (same industry clustering of formal or informal sectors) and co-location (clustering of formal sector with informal sector or other ways) on firm-level productivity.

Further, Chinitz and Jacobs’ urbanisation index⁵ analyses the diversity of industries in a region. A diversified region in terms of industries may facilitate the transfer of new knowledge, help to access business services, in turn, reduce the cost of firms and enhance the productivity. On the other hand, Moreno-Monroy (2012) shows that extreme cost competition may offset the scope of knowledge or technological spillover when firms produce in a region with many other enterprises. Therefore, it can be argued that diversification may generate positive or negative externalities. We make use of the Herfindahl index to measure the degree of diversification of a region, which measures the sum of the squares of the employment shares of all the industries in a region r. The index is given as follows:

U_{r} = \sum_{j = 1}^{n} {(E_{j r} / E_{r})}^{2}

(2)

where E_jr is the employment in industry j in region r, and E_r is the total employment in region r irrespective of industries. Similar to the localisation index, we distinguish the diversification index as follows: formal sector diversification—measures the diversity in the formal sector activity in a region and informal sector diversification—accounts for the diversity in the informal sector activity in a region. Further, we include a host of control variables in the estimation, including market size (defined as the log of population). Since the location of firms influence productivity, we include a dummy variable for the urban region (which takes the value 1, and 0 otherwise). In the case of an informal firm productivity, we control for the size of the enterprise by including dummy for non-directory manufacturing enterprises (NDMEs) and own-account manufacturing enterprises (OAMEs).⁶

IV. Methodology

We estimate the following Equation (3) as follows:

Log {(LP)}_{i} = \propto_{1} + β_{1} X_{i} + β_{2} L_{j r} + β_{3} U_{r} + β_{4} X_{r} + \dots + e_{i}

(3)

where, Log(LP) _i denotes log of labour productivity of each plant. X_i is the observed characteristics of the plant (share of skilled labour in the case of formal sector and share of wage workers in the case of informal sector). L_jr is the vector of variables denoting localisation economies (includes both the formal localisation and informal localisation). Similarly, U_r denotes the vector of urbanisation economies’ variables (includes both formal urbanisation and informal urbanisation, measured using the urbanisation index given in Equation (2. X_r denotes the control for enterprise size and region size. e_i is the random error term. We estimate this equation separately for the formal sector and informal sector.

There are certain econometric issues while analysing the importance of agglomeration economies on plant productivity. These include the possibility of endogeneity and selection bias. If a region is endowed with the conditions that favour higher productivity, leading to higher concentration of plants and workers, then that region becomes larger (Brulhart, 1998; Moomaw, 1983). Further, more productive firms may have a tendency to self-select into productive clusters (Howard et al., 2014). These problems may cause the correlation across regions in the error term in the estimation. Ciccone and Hall (1996) addressed the endogeneity issue by employing the historical population and railroads as the instruments in the analysis in the case of the USA. Similarly, Tran and La (2018) addressed the endogeneity issue by using lagged effects of agglomeration variables and historical population variables. Following these studies, in our empirical analysis, we use the lagged effects of each agglomeration variables⁷ and district-level population of 2001 census as the instruments. Therefore, we estimate Equation (3) using the instrumental variable (IV) regression.

To overcome the issue of self-selection bias, we include one observed characteristic of the enterprise in the model. To analyse the impact of agglomeration economies on formal sector productivity, we use skilled workers as the observed characteristic (defined as the proportion of the managerial and supervisory employees to the total employees). We assume that those plants with a higher share of skilled labour will be relatively more productive. Similarly, in the case of the informal sector, we use the proportion of wage workers to the total workers in the enterprise. In the case of India, more than 60% of informal sector enterprises fall under the category of OAMEs, which use unpaid family members in their production. Therefore, those plants that employ wage workers will be more productive (McCaig & Pavcnik, 2014). Therefore, we assume that firms with more skilled workers in the formal sector and more wage workers in the informal sector are more productive. Details of the variables and summary statistics are presented in Table 1.

Table 1.

Variable Description and Summary Statistics.

Variables	Description	Varies By	Formal Sector		Informal Sector		Source
Variables	Description	Varies By	Mean	SD	Mean	SD	Source
Log_labour productivity (logLP)	Log (total output/total number of workers)	Plant	13.259	1.299	7.803	0.896	ASI, NSSO
Share of skilled workers	Proportion of skilled workers to the total number of workers	Plant	0.157	0.231			ASI
Share of waged workers	Proportion of hired workers to the total number of workers	Plant			0.135	0.258	NSSO
Formal localisation 2010	Rate of concentration of an industry in a region compared to the country in formal sector in 2010	Industry × region	0.504	1.241	0.21	2.448	ASI, NSSO
Informal localisation 2010	Rate of concentration of an industry in a region compared to the country in informal sector in 2010	Industry × region	0.353	2.742	0.688	9.344	ASI, NSSO
Formal urbanisation 2010	Diversity of (formal) industries in a region in 2010	Region	0.179	0.166	0.275	0.095	ASI, NSSO
Informal urbanisation 2010	Diversity of (informal) industries in a region in 2010	Region	0.277	0.093	0.237	0.205	ASI, NSSO
Formal localisation 2006	Rate of concentration of an industry in a region compared to the country in formal sector in 2006	Industry × region	0.402	1.589	0.181	2.057	ASI, NSSO
Informal localisation 2006	Rate of concentration of an industry in a region compared to the country in informal sector in 2006	Industry × region	0.549	1.693	1.245	11.78	ASI, NSSO
Formal urbanisation 2006	Diversity of (formal)industries in a region in 2006	Region	0.079	0.104	0.014	0.016	ASI, NSSO
Informal urbanisation 2006	Diversity of (informal)industries in a region in 2006	Region	0.013	0.011	0.108	0.141	ASI, NSSO
Market size	log (population in a region)	Region	14.69	0.801			Census
Enterprise size	Dummy Variable (NDMEs = 1, OAMEs = 0)	Plant			1.23	0.42	NSSO
Region	Dummy Variable (Rural = 0, Urban = 1)	Plant	1.59	0.491	1.464	0.498	ASI, NSSO

Source: Authors’ compilation.

Note: ASI: Annual Survey of Industries; NSSO: National Sample Survey Office.

V. Results

The empirical analysis includes the agglomeration effects on productivity of both formal and informal sectors. First, we consider the productivity change in the formal manufacturing plants by their localisation and by the co-location with informal manufacturing plants. The results are presented in Table 2.

Table 2.

Formal Firm Productivity and Agglomeration Economies.

Variables	(1) logLP	(2) logLP	(3) logLP
Formal localisation_2010	0.177** (0.0875)		0.256** (0.101)
Informal localisation_2010	0.227*** (0.0224)		0.227*** (0.0254)
Formal urbanisation_2010		–5.280*** (0.927)	–7.551*** (0.940)
Informal urbanisation_2010		0.575 (2.612)	–4.923*** (1.862)
Skilled workers	1.192*** (0.0596)	0.927*** (0.0641)	0.801*** (0.0846)
Market size	0.140*** (0.0176)	–0.114*** (0.0421)	–0.162*** (0.0419)
Region size	0.350*** (0.0282)	0.167*** (0.0293)	0.184*** (0.0395)
Constant	10.64*** (0.267)	15.47*** (1.445)	17.92*** (1.156)
Observations	26157	26157	26157
Cragg–Donald Wald F statistic	77.91	12.03	12.28
Sargan statistic (p-value)			0.207 (0.648)

Source: Authors’ compilation.

Note: Standard errors in parentheses. *** denotes p < 0.01, ** denotes p < 0.05 and * denotes p < 0.1. Columns 1 and 2 are exactly identified.

In the analysis, we first observe the impact of localisation of both formal and informal sectors on the firm’s productivity. Here, the coefficient value of formal localisation shows the productivity change in a firm due to its choice to locate near to the same industry. On the other hand, coefficient value of informal localisation shows the impact of the co-location of formal and informal sectors on formal plant productivity. Further, in the second column, we include another agglomeration economies measure, that is, Chinitz and Jacobs urbanisation, which shows how much the region is diversified in terms of its manufacturing activity. Similar to localisation, we include the index in terms of both formal and informal sector. Finally, we include all the agglomeration variables together and analyse its impact on formal and informal plant-level productivity.

The estimates in Table 2 show that localisation of the formal sector or same industry clustering generate a positive externality on formal sector plants, which lend support to the arguments put forward by Marshall (1920). It is mainly due to the clustering of different plants in the same industry, as they can take advantage of knowledge spillovers, sharing of inputs and skilled labour. This finding is in line with the earlier empirical evidence (Duranton & Puga, 2001; Krugman, 1991; Rosenthal & Strange, 2004). Similarly, localisation of the informal sector generates positive productivity effects to formal manufacturing. Therefore, we find that plants belonging to the formal sector benefit from collocating with the informal sector. This result supports the argument of input–output linkages and subcontracting between formal and informal sectors (Mukim, 2015). Further, the urbanisation index of both formal and informal sectors, which shows clustering of plants with other industries, brings negative externality. This result explains that a higher rate of diversification in a region may lead to the problem of congestion, which, in turn, adversely affects the plant productivity in the region. Similarly, Ghani et al. (2012) argue that high population density and strict labour regulations discourage firms to locate in urbanised regions, which support the previous findings of Overman and Venables (2010).

Along with these agglomeration variables, to reduce the problem of selection bias, we include ‘share of skilled workers’ in the estimation. The coefficient of this variable is positive, which supports our hypothesis that those plants which employ more skilled workers are more productive. Further, when we analyse the market size, measured using the population data, we observe an adverse impact on plant-level productivity. The coefficient of the regional dummy (rural or urban) reveals that plants in urban regions are more productive.

To resolve the problem of endogeneity, following Tran and La (2018), and Ciccone and Hall (1996), we include the lagged effects of agglomeration variables and population data as IVs. Since the observation year is 2010–2011, the IVs include all formal and informal localisation and urbanisation indices in 2005–2006 and district population density of 2001. To ensure the validity of our IVs, we perform certain post-estimation tests, which include weak identification test and Sargan statistic for over-identification. The IV post estimation supports the selection of IVs that at least one instrument for all endogenous variables is statistically significant.

We carried out similar analysis for the informal sector to tease out the effects of agglomeration economies on the productivity of informal enterprises, presented in Table 3. In the case of informal sector, also, we find Marshallian agglomeration economies, that is, positive productivity effects through the localisation of same industry sector (informal localisation). The coefficient of formal sector localisation turns out to be insignificant, possibly indicating that informal enterprises are unable to enjoy productivity benefits by locating near formal firms belonging to the same industry. This may be because of the congestion generated through co-location within the same industry (Tran & La, 2018). Interestingly, diversified economic activity in a region in terms of both formal and informal activities (formal urbanisation and informal urbanisation) show a positive value. This set of results show that informal firms enjoy positive productivity benefits by locating themselves in a diverse region. Therefore, this result lends support to the argument to the movement of firms, that is, informal firms migrate into urbanised areas in search of benefits generated from a diversified region (Ghani & Kanbur, 2013).

Table 3.

Informal Firm’s Productivity and Agglomeration Economies.

Variables	(1) Loglp	(2) Loglp	(3) Loglp
Informal localisation_2011	0.00829*** (0.00209)		0.0153*** (0.00359)
Formal localisation_2011	0.00310*** (0.000668)		–0.00128 (0.00124)
Informal urbanisation_2011		6.671*** (0.754)	6.101*** (0.561)
Formal urbanisation_2011		1.917*** (0.261)	1.691*** (0.199)
Waged workers	0.520*** (0.0528)	0.528*** (0.0962)	0.508*** (0.0889)
Enterprise size	0.350*** (0.0323)	0.459*** (0.0603)	0.455*** (0.0552)
Region size	0.197*** (0.00947)	0.129*** (0.0207)	0.135*** (0.0188)
Constant	7.557*** (0.00678)	3.340*** (0.486)	3.799*** (0.342)
Observations	62801	62801	62801
Cragg–Donald Wald F statistic	2100	51.335	35.128
Sargen statistic (p-value)			2.218 (0.1364)

Source: Authors’ compilation.

Note: (a) Standard errors in parentheses. (b) *** p < 0.01, ** p < 0.05 and * p < 0.1. (c) Columns 1 and 2 are exactly identified.

In the case of informal sector, to reduce the problem of selection bias, we include ‘wage labour’ in the analysis. The positive coefficient value of wage labour stays in line with the findings of Tran and La (2018). Similarly, we include enterprise size dummy (NDMEs and OAMEs), which indicates NDMEs are more productive as compared to OAMEs. Further, we include the regional dummy (rural and urban), which shows that enterprises in urban areas are more productive than the enterprises located in rural areas. All the post-estimation tests of IV regression support validity of our IVs.

VI. Conclusion

The objective of this study is to examine the co-agglomeration of formal and informal manufacturing sectors and plant-level productivity. Empirical results of the study show that formal firms benefit from the externalities generated from the formal sector localisation and the informal sector localisation, indicating that clustering with the same industry irrespective of formal or informal sectors helps the formal firms to enhance their productivity. Further, urbanisation indices, which imply the diversification of region, adversely affect the productivity. This result suggests that the increases in congestion costs over the benefits of agglomeration may lead to adverse effect.

Similar to formal firms, informal firms benefit from the clustering of informal firms within industries (informal localisation). However, their co-location with formal firms within industries show insignificant effect. Further, the location of informal firms in diversified (in terms of both formal and informal sector activities) regions helps them to improve productivity benefits. The result suggests that informal firms may benefit from this diversity of the region and the agglomeration externalities it creates. Therefore, our results support the observation of the large movement of informal firms to urban areas and indicate the importance of the informal sector in the rapid urbanisation in developing economies.

Footnotes

Declaration of Conflicting Interests

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Notes

ORCID iD

Renjith Ramachandran

References

Abraham

(2017). Informality in the Indian labour market: An analysis of forms and determinants. The Indian Journal of Labour Economics , 60(2), 191–215.

Arimah

B. C.

(2001). Nature and determinants of the linkages between informal and formal sector enterprises in Nigeria. African Development Review , 13(1), 114–144.

Black

, Henderson

(2003). Urban evolution in the USA. Journal of Economic Geography , 3(4), 343–372.

Brulhart

(1998). Economic geography, industry location and trade: the evidence. The World Economy , 21(6), 775–801.

Ciccone

, Hall

R. E.

(1996). Productivity and the density of economic activity. American Economic Review , 86, 54–70.

Duranton

(2008). From cities to productivity and growth in developing countries. Canadian Journal of Economics/Revue canadienne d'économique , 41(3), 689–736.

Duranton

, Puga

(2001). Nursery cities: Urban diversity, process innovation, and the life cycle of products. American Economic Review , 91(5), 1454–1477.

Ghani

, Kanbur

(2013). Urbanization and (in) formalization . The World Bank.

Ghani

, Goswami

A. G.

, Kerr

W. R.

(2012). Is India's manufacturing sector moving away from cities? The World Bank.

10.

Henderson

J. V.

(2003). Marshall's scale economies. Journal of Urban Economics , 53(1), 1–28.

11.

Howard

, Newman

, Rand

, & Tarp

(2014). Productivity-enhancing manufacturing clusters: Evidence from Vietnam (No. 2014/071). WIDER Working Paper.

12.

Jacobs

(1969). The economy of cities . Random House.

13.

Jütting

, & de Laiglesia

J. R.

(Eds.). (2009). Is informal normal? Towards more and better jobs in developing countries . Development Centre of the Organisation for Economic Co-operation and Development.

14.

Krugman

(1991). Increasing returns and economic geography. Journal of political economy , 99(3), 483–499.

15.

Lall

S. V.

, Shalizi

, & Diechmann

(2004). Agglomeration economies and productivity in Indian industry. Journal of Development Economics , 73, 643–673.

16.

Maré

, Timmins

(2006). Geographic concentration and firm productivity (No. 1124-2019-3307). Ministry of Economic Development.

17.

Marshall

(1920). Principles of economics (Vol. 8). Macmillan and Co Ltd.

18.

Martin

L. A.

, Nataraj

, Harrison

A. E.

(2017). In with the big, out with the small: Removing small-scale reservations in India. American Economic Review , 107(2), 354–86.

19.

McCaig

, Pavcnik

(2018). Export markets and labor allocation in a low-income country. American Economic Review , 108(7), 1899–1941.

20.

Moomaw

R. L.

(1983). Production efficiency and region. Southern Economic Journal , 48, 344–357.

21.

Moreno-Monroy

(2012). Critical commentary. Informality in space: Understanding agglomeration economies during economic development. Urban Studies , 49(10), 2019–2030.

22.

Mukim

(2015). Coagglomeration of formal and informal industry: Evidence from India. Journal of Economic Geography , 15(2), 329–351.

23.

Overman

H. G.

, Venables

A. J.

(2005). Cities in the developing world (No. 695). Centre for Economic Performance, London School of Economics and Political Science.

24.

Overman

H. G.

, Rice

, Venables

A. J.

(2010). Economic linkages across space. Regional Studies , 44(1), 17–33.

25.

Ranis

, Stewart

(1999). V-goods and the role of the urban informal sector in development. Economic Development and Cultural Change , 47(2), 259–288.

26.

Rosenthal

S. S.

, & Strange

W. C.

(2004). Evidence on the nature and sources of agglomeration economies. In Handbook of regional and urban economics (Vol. 4, pp. 2119–2171). Elsevier.

27.

Tran

T. B.

, La

H. A.

(2018). Agglomeration effects: Productivity of the informal sector in Vietnam. The Journal of Development Studies , 54(2), 292–311.

28.

Venables

A. J.

(1996). Equilibrium locations of vertically linked industries. International Economic Review , 37(2), 341–359.