ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Temporary and Seasonal Migration: Regional Pattern, Characteristics and Associated Factors

The regional pattern of temporary and seasonal labour migration in India assumes sharp focus when seen in the light of data from the 64th round of the National Sample Survey. The phenomenon is more prevalent in rural areas of the country's northern and eastern states. This paper also examines the association between temporary migration and its determining factors, particularly economic status, landholding and educational levels. It observes that there is a significant negative association between economic and educational attainment and temporary migration, both in rural and urban areas. In general, socio-economically deprived groups such as adivasis and those from the lower castes have a greater propensity to migrate seasonally, which also reflects its distress-driven nature.


Temporary and Seasonal Migration: Regional Pattern, Characteristics and Associated Factors

Kunal Keshri, R B Bhagat

The regional pattern of temporary and seasonal labour migration in India assumes sharp focus when seen in the light of data from the 64th round of the National Sample Survey. The phenomenon is more prevalent in rural areas of the country’s northern and eastern states. This paper also examines the association between temporary migration and its determining factors, particularly economic status, landholding and educational levels. It observes that there is a significant negative association between economic and educational attainment and temporary migration, both in rural and urban areas. In general, socio-economically deprived groups such as adivasis and those from the lower castes have a greater propensity to migrate seasonally, which also reflects its distress-driven nature.

An earlier version of the paper was presented at an “International Conference on Population Dynamism of Asia: Issues and Challenges Ahead” during 11-13 July 2011 at the Department of Geography, University of Malaya, Kuala Lumpur, Malaysia. The authors are thankful to the participants for their feedback and suggestions, and they would like to acknowledge Rajesh Kumar Chauhan and Gopal Agrawal for their help in extraction of NSS data and Sanjay Mohanty and Kirti Gaur for their useful comments towards the improvement of the manuscript.

Kunal Keshri ( and R B Bhagat (rbbhagat@ are with the International Institute for Population Sciences, Mumbai.

1 Introduction

emporary migration, often used interchangeably with circular, seasonal, short-term and spontaneous migration, has been a subject of much discourse. According to Zelinsky (1971), all these movements, usually short-term, repetitive or cyclic, having the common motive of a temporary change of residence, are circular in nature. Circular migrants follow a circular path and maintain continuous but temporary absences from their place of origin for more than one day (Hugo 1982). Temporary or circular migration is a move made for a short period of time with the intention of returning to the place of usual residence. An important group of temporary migrants consists of seasonal migrants, who combine activity at several places according to seasonal labour requirements (Keshri and Bhagat 2010).

Prevailing regional inequalities and uneven development in many Asian countries impel temporary internal migration from agriculturally backward and poor rural areas. Temporary migration has increased substantially in the last two decades in south, south-east and east Asia (Brauw 2007; Deshingkar and Akter 2009; Deshingkar and Grimm 2005; Ha et al 2009; Lam et al 2007). Seasonal migration has long been a source of income for rural households unable to support themselves through agriculture. Households diversify their economic activities outside the traditional agricultural sphere by sending out members to work in urban areas in the lean period (Pham and Hill 2008). According to the school of New Economics of Labour Migration (NELM), temporary migration is considered a risk diversification strategy (Prothero and Chapman 1985; Stark and Bloom 1985; Stark and Levhari 1982).

It is evident from the extant literature that temporary migration is one of the most significant livelihood strategies adopted by the poorest sections in rural India, predominantly in the form of seasonal mobility of labour (Breman 1978, 1996; Deshingkar and Farrington 2009; Deshingkar and Start 2003; Haberfeld et al 1999; Mosse et al 2005; Rao and Rana 1997; Rogaly 1998; Rogaly et al 2001; Srivastava and Sasikumar 2003). People also move from rural areas to nearby or distant cities to find jobs in construction or the unorganised informal sector (Breman 1994; Deshingkar and Farrington 2009; Haberfeld et al 1999; Vijay 2005). Mukherji (2006) has termed this distress migration, which, according to him, paves the way for urban decay by causing urban poverty, unemployment and a shortage of housing. Breman (1994), on the other hand, sees seasonal labour migration in western India as an important survival option for landless labourers. Landless

Economic & Political Weekly

january 28, 2012 vol xlvii no 4

agricultural labourers in Gujarat, Bihar, Madhya Pradesh, West Bengal and Jharkhand, who are trapped in debt bondage and belong to the lower social strata (scheduled tribes and castes or STs and SCs), migrate seasonally within or outside their states (Breman 1994; Deshingkar and Farrington 2009; Haberfeld et al 1999; Jayaraman 1979; Rogaly 1998; Rogaly et al 2001; Vijay 2005). For instance, the monsoon frequently fails in Panchmahals district of Gujarat and seasonal migration of the tribal population to nearby rural and urban areas is common (Jayaraman 1979). Similar circumstances prompt temporary migration among tribal women in Jharkhand and West Bengal as well (Dayal and Karan 2003; Rogaly et al 2001). Though such migration can be taken as a sign of dynamism, it has more to do with increasing inequalities, agrarian instability and inadequate livelihood generation in many parts of rural and urban India (Chandrasekhar and Ghosh 2007; Keshri and Bhagat 2010).

There are several demographic and socio-economic factors such as age, sex, educational attainment, social group or caste, religion, poverty and size of landholding that affect temporary migration (Brauw 2007; Deshingkar and Grimm 2005; Deshingkar 2006; Ha et al 2009; Lam et al 2007; Pham and Hill 2008). Yang and Guo (1999) have found that in rural areas, the decision of men to migrate is mainly moulded by community-level factors, while among women, temporary labour migration is predominantly determined by individual characteristics. Among the broad group of the underclass or the socio-economically deprived, which includes the poorest of the poor, the landless, illiterates or those with a very low level of education (say, primary school), the SC/STs and Muslims, temporary migration is very high (Bird and Deshingkar 2009; Connell et al 1976; Dayal and Karan 2003; Deshingkar 2006; Hugo 1985; Mosse et al 2005; Vanwey 2003). Poverty is supposed to be a key push factor in temporary migration. Skeldon (2002) states that under certain conditions, poverty may be the root cause of migration in some parts of the world, whereas in other parts, under different conditions, the poor may be among the last to move. Brauw (2007) finds that households having low annual expenditures are more prone to migrate than others.

Some studies using data from the National Sample Survey (NSS) and Census of India have established that poor people are less mobile as far as permanent or semi-permanent migration is concerned (Bhagat 2010; Singh 2009), while Kundu and Sarangi (2007) find there is no association between poverty and seasonal migration across the urban centres. Recent work by Keshri and Bhagat (2010), which utilises data from the 55th round of the NSS, reveals that seasonal migration is very prevalent among those belonging to the lowest expenditure quintiles, rural areas and STs. States having a higher level of inequality show higher temporary migration rates. However, the data has some limitations because the sample of temporary migrants is small and information is lacking on their destinations and occupations.

Despite large-scale temporary migration in absolute numbers, the phenomenon has not been adequately studied at the macro level in India. This has possibly been due to the unavailability of national-level data or the very limited information collected by national surveys (as in the 55th round of the NSS). The census, which is the other important source of migration data, is mainly concerned with current and permanent migration and does not attempt to capture seasonal or short-term flows of labour (Chandrasekhar and Ghosh 2007; Keshri and Bhagat 2010). There is thus a dearth of studies that provide a general picture of temporary migration at the national and state levels, which also examine its determining factors.

Against this backdrop, the recently available data from the 64th round of the NSS (2007-08) provides a great opportunity to study temporary migration in India. With a comparatively large sample size of temporary migrants, this data allows us to analyse the phenomenon at the state level. Moreover, there is information on the destinations of temporary migrants, which makes studying streams of migration possible. The quality of the data has also improved compared to the previous round.1 Therefore this study aims to explore the pattern and state-wise intensity of temporary migration and to examine its association with poverty, landholding and education after controlling for other sociodemographic factors. As temporary migration in India is strongly influenced by seasons, the terms “temporary” and “seasonal” are used interchangeably. In this study, a temporary migrant is defined as a household member who has stayed away from his or her village or town for one month or more but less than six months in the last 365 days for employment or in search of employment.

2 Data

The present study utilises unit level data from the 64th round of the NSS, which is a large-scale, nationally representative, multiround survey (NSSO 2010). In this round, information on various facets of migration was collected through a schedule on “Employment and Unemployment and Migration Particulars” (Schedule 10.2). Two blocks were canvassed for the particulars and the survey was conducted in all states and union territories from 1 July 2007 to 30 June 2008 (the reference period of the survey). For Schedule 10.2 at the all-India level, 12,688 first stage units (7,984 villages and 4,704 urban blocks) were covered. In the central sample, the survey covered a sample of 1,25,578 households (79,091 in rural areas and 46,487 in urban areas) and a sample of 5,72,254 persons (3,74,294 in rural areas and 1,97,960 in urban areas).

In this round of the NSS, voluminous data on short-duration movements or temporary migration was collected. Information about temporary migrants was collected by asking heads of households whether any member had stayed away from the village or town for a period of 30 days to six months during the last 365 days in search of employment or for employment. Apart from this, information on the number of spells (staying away from a village or town for 15 days or more was termed a spell), destination stayed at during the longest spell (such as, the same district, the same state but another district, another state, another country, and so on) and if worked, the industry worked in were collected. In the previous migration-related version (55th round, 1999-2000) of this survey, an attempt was made to identify temporary migrants by considering people who had stayed away from their villages or towns for 60 days or more for employment or in search of employment (NSSO 2001). But in the 64th round of

january 28, 2012 vol xlvii no 4


the survey the minimum number of days was lowered to 30 days to capture short-term movements more effectively.

The usual place of residence was defined in the NSS as a place (village or town) where a person has stayed continuously for six months or more. If a household member’s last usual place of residence, anytime in the past, was different from the place of enumeration, he or she was considered a migrant. The NSS used a stratified multistage sampling design and appropriate multipliers and weights have been used to generate national and state-level estimates. Details of the multipliers and sampling weights used are in the NSS report pertaining to migration (NSSO 2010).

3 Analytical Strategy and Cataloguing of Variables

The rate of temporary migration was calculated to study the pattern and intensity of migration. The temporary migration rate for a category of persons or a region (say, district or state) for a specified period of time was estimated by the number of migrants of that category per 1,000 persons of that category in the region. It was calculated using the following formula:

Total number of migrants in a particular category Temporary migration rate = × 1,000

Total number of persons in


that category

To study the regional pattern of temporary migration, statewise estimates were generated. The sample was restricted to the working-age population (15 to 65 years) for all bi-variate and multivariate analyses since temporary migration is mainly for employment (Keshri and Bhagat 2010; Yang and Guo 1999).2 However, state-wise estimates were generated for the workingage population as well as for all ages. Using the information on destination during the longest spell, streams of migration were identified. The variable for monthly per capita consumer expenditure quintiles (MPCE quintiles) was obtained by dividing the total household expenditure by the household size and then distributing households into five equal percentile groups. The quintiles were defined as lowest, lower, medium, higher and highest.

Binary logistic regression models were fitted to assess the adjusted effects of socio-economic characteristics on the likelihood of a person being a seasonal migrant.3 The outcome variable of seasonal migration was coded in a binary form, that is, “1” if a person was a temporary migrant and “0” if not. In the absence of income-related data in Indian sample surveys, the household consumer expenditure data of the NSS provides quite a close view of the economic conditions of households. Therefore, MPCE quintiles were taken as a proxy of the economic condition of households, as many recent studies have done (Banerjee and Raju 2009; Keshri and Bhagat 2010; Kundu and Sarangi 2007). This was categorised into low (reference), medium and high for logistic regression analysis.

Connell et al (1976) argue that landholding is the primary economic force that drives temporary migration in rural India. It is established that temporary migration is the main source of income for rural households who are unable to support themselves through agriculture in their home communities, particularly those that have small landholdings or are landless (Hugo 1985; Vanwey 2003). Variable landholding was included in the models

Economic & Political Weekly

january 28, 2012 vol xlvii no 4

with less than 1 hectare as the reference category. Educational attainment is the other important indicator of socio-economic development that has been found to be associated with temporary migration (Yang 1992). It was classified into four broad categories – below primary (reference), middle level completed, and secondary or higher educated (graduate and above).

In India, there is economic stratification of the population in accordance with various social groups and castes and this is related to migration, particularly in rural areas, as many earlier studies have found (Gnanou 2008; Haberfeld et al 1999; Keshri and Bhagat 2010; Mosse et al 2005; Vijay 2005). For a logistic regression analysis, the social group STs was assigned as the reference category in a threefold classification of caste – STs, SCs, Other Backward Classes (OBCs) and others. Religion was also taken as a control variable with the categories Hindu (reference), Muslim and Others. The last category consisted of all other religious groups. For the variable sex, male was taken as the reference group. Marital status was also included in the model (with single as the reference category). To know the likelihood of temporary migration, the variable state/union territory was included in the models. Rajasthan was taken as the reference category since its seasonal migration rates in rural and urban areas are equal to the national average.

Four different sets of logistic regression models were applied for examining the associated factors of temporary migration. In model I, for which regression was run on the sample of the workingage population of rural areas, MPCE, landholding and educational attainment were taken into consideration. In model II, along with MPCE quintiles, landholding and education, caste, religion, household size, sex, marital status and state were included as control variables. In model III, regression was run for the urban sample using the same variables as those in model I. In model IV, the same set of variables as model II were included using the urban sample to assess the independent impact of these factors on temporary migration. For ease of interpretation of the result, only those states that were statistically significant were shown in the logistic regression tables.

4 Results and Discussion
4.1 Regional Pattern of Temporary and Seasonal Migration

Table 1 (p 84) presents the regional picture of temporary and seasonal migration in India by showing the estimated number of temporary migrants and temporary migration rate (migrants per 1,000 population) across states for all ages and in the workingage population (15-64 years).

There were a total of 1,36,21,100 temporary migrants of all ages in India in the reference period 2007-08. This figure falls to 1,30,76,500 when only the working-age population is taken into account. Among the major states, the estimated number of temporary and seasonal migrants in the working-age population was highest (20,85,600) in Bihar, followed by Uttar Pradesh (18,96,500), West Bengal (15,28,400) and Madhya Pradesh (12,36,900). Migration rates were calculated to assess the intensity of migration, which was found to be highest in Bihar (50 per 1,000).

25 52 29 232 5 5 5

Table 1: Temporary and Seasonal Migrants, Temporary and Seasonal Migration Rate (Migrants Per Thousand), National Sample Survey, 2007-08

States Temporary and Seasonal Migrants Temporary and Seasonal (in Thousands) Migration Rate (per 1,000) All Ages Age Group 15-64 Years All Ages Age Group 15-64 Years

Andhra Pradesh 789.5 725.8 10.5 14.2

Arunachal Pradesh 18.9 17.8 17.6 25.8

Assam 294.3 286.5 11.8 17.4

Bihar 2,125.7 2,085.6 28.2 49.9

Chhattisgarh 329.7 262.7 14.3 18.3

Delhi 52.1 52.1 4.0 5.8

Goa 8.2 8.1 5.7 8.0

Gujarat 1,147.8 1,107.9 23.2 33.8

Haryana 72.7 69.2 3.3 5.0

Himachal Pradesh 29.6 29.4 4.7 7.2

Jammu and Kashmir 102.7 101.9 12.4 18.6

Jharkhand 539.6 530.8 21.9 35.9

Karnataka 437.9 420.5 8.9 12.7

Kerala 132.0 127.2 4.4 6.3

Madhya Pradesh 1,262.0 1,236.9 20.9 33.5

Maharashtra 728.1 682.1 7.7 10.7

Manipur 8.6 6.8 4.3 5.1

Meghalaya 25.8 25.7 11.1 18.0

Mizoram 4.4 4.1 5.0 7.0

Nagaland 32.8 32.6 34.0 48.5

Orissa 437.8 430.3 12.0 18.2

Punjab 127.7 89.8 5.4 5.6

Rajasthan 737.3 724.0 12.7 21.1

Sikkim 1.7 1.3 3.3 3.9

Tamil Nadu 578.6 542.4 9.4 12.8

Tripura 12.8 11.2 3.7 4.8

Uttar Pradesh 1,974.4 1,896.5 11.6 19.6

Uttarakhand 30.9 30.5 3.6 5.7

West Bengal 1,569.4 1,528.4 20.0 29.3

India 13,621.1 13,076.5 13.5 20.5

Total of migrants of all states may not equal to the all-India figure because it also includes the migrants from union territories. Source: 64th National Sample Survey 2007-08, unit level data.

Among the major states, Jharkhand (36), Gujarat (34), Madhya Pradesh (33), West Bengal (30) and Rajasthan (21) also had a high intensity of seasonal migration above the national average (20). Uttar Pradesh, Jammu and Kashmir, Chhattisgarh, and Orissa showed moderate levels (15 to 20 per 1,000) of seasonal migration. The remaining states had a low intensity of migration. Among the smaller states, Nagaland had the highest migration rate (48) while it was insignificant in the remaining north-eastern states. Differentials across states were more or less similar when we consider migrants of all ages. Most of the states had high levels of intra-regional inequality and seasonal migration was an important livelihood strategy during the agricultural lean season with people moving to big cities in the same state or to other states (Breman 1994; Deshingkar and Farrington 2009; Jayaraman 1979; Rogaly et al 2001).

Temporary migration in rural and urban areas had different driving forces due to differences in the socio-economic status of the populations. Therefore it is imperative to analyse the phenomena separately. Figure 1 shows the temporary migration rate by place of residence across states. We find a significant difference between the temporary migration rates of rural (26) and urban areas (6). In rural areas, Bihar had the highest seasonal

Figure 1: Rural-Urban Differentials in Temporary and Seasonal Migration Rate (Migrants Per Thousand, Age-Group 15-64 Years), Indian States, National Sample

Survey, 2007-08
6 26
West Bengal
8 37
Uttarakhand 2 7
Uttar Pradesh 5 24
Tripura 4 10
Tamil Nadu 8 16
Sikkim 0 5
6 26
3 7
Orissa 6 20
Nagaland 43 66
Mizoram 6 8
Meghalaya 5 21
5 5
Maharashtra 2 17
Madhya Pradesh 8 42
Kerala 4 7
Karnataka 6 16
Jharkhand 1 44
Jammu and Kashmir
10 21
Himachal Pradesh 2 8
Haryana 2 6
Gujarat 7 51
4 13
Delhi 3 6
Chhattisgarh 4 21
12 54
18 Urban
Assam 17
Arunachal Pradesh 26
Andhra Pradesh
2 19 Rural
0 20 40 60

Migration rate (per 1,000) Source: 64th National Sample Survey 2007-08, unit level data.

Figure 2: Streams of Temporary and Seasonal Migration in India according to Sex (Age-Group 15-64 Years), National Sample Survey, 2007-08

Urban to urban Urban to rural

Rural to rural

Rural to urban


Male Female Total Per cent distribution of migrants Source: 64th National Sample Survey 2007-08, unit level data.

migration rate (54), followed by Gujarat (51), Jharkhand (44), Madhya Pradesh (42) and West Bengal (37). In urban areas, Assam had a high temporary migration rate (18), along with Bihar (12) and Jammu and Kashmir (10). In general, rural areas had very high migration rates compared to urban areas. In Jharkhand, Gujarat, Chhattisgarh, Andhra Pradesh, Maharashtra, West Bengal,

january 28, 2012 vol xlvii no 4


Uttar Pradesh, Madhya Pradesh and Rajasthan the migration rate in rural areas was many times higher than that in urban areas. Jammu and Kashmir, Orissa, Karnataka, Punjab and Uttarakhand also showed comparatively higher temporary migration rates in rural than urban areas. Interestingly, in states such as Assam and Nagaland, the urban temporary migration rate was higher than the rural. The findings show that rural-urban differences in migration rates were not very significant in the southern and north-eastern states.

Figure 2 (p 84) shows the distribution (percentage) of male and female temporary and seasonal migrants according to the four migration streams in India. It is seen that overall more than half the migrants were in the rural to urban stream, followed by the rural to rural stream. However, among females, the rural to rural migration rate was higher. The dominance of rural to urban migration among males reflected the increasing differences between rural and urban areas in India in terms of income and employment. The informal sector in urban areas attracts poor people from rural areas, mainly when there is a lull in agricultural work.

4.2 Characteristics of Temporary and Seasonal Migration

The relationship between poverty and migration has long been a subject of debate. It is well recognised that poor people migrate for survival within the country and this mobility is generally in the form of short-term migration, even though the capacity to afford migration is low among the poor (Kundu and Sarangi 2007; Skeldon 2002). In the absence of a direct measurement of poverty, MPCE quintiles have been used as indicators to unravel whether temporary mobility

Table 2: Temporary Migration Rate is higher among the poor or (Migrants Per Thousand) according to

MPCE by Place of Residence (Age-Group 15

rich. Table 2 shows the cross

64 Years), National Sample Survey, 2007-08

classification of migration

MPCE Quintiles Rural Urban Total

rates according to MPCE quin- 44.8 40.6


tiles by place of residence. Lower 32.1 6.2 25.6

The result showed that the Medium 23.8 4.6 17.0

temporary migration rate Higher 17.4 5.0 10.2

was very high (40) among Highest 11.3 2.3 5.8 Total 26.4 5.5 20.5

those in the lowest MPCE

Source: Same as in Table 1.

quintile and it decreased in the higher quintiles. Among those in the lowest quintile in rural areas, the temporary migration rate was almost 45 per 1,000 while it was 32 per 1,000 in the lower quintile and further decreased with increasing MPCE. The trend in urban areas was similar but the degree of Table 3: Temporary and Seasonal Migration

Rate (Migrants Per Thousand) according to

change was smoother from

Land Possession by Place of Residence the lowest to the highest quin-(Age-Group 15-64 Years), National Sample

Survey, 2007-08

tile. Nonetheless, we find re-

Landholding (in hectares) Rural Urban Total

markable rural-urban differ-

Less than 1 29.1 5.4 21.2

entials with respect to MPCE

1-4 18.4 7.0 17.8 quintiles in migration rates. More than 4 10.9 11.9 10.9

On the whole, seasonal mi-Total 26.4 5.5 20.5

Source: Same as in Table 1.

gration rates fell with an

increase in the size of land possessed by households (Table 3). In rural areas, those belonging to households having less than 1 hectare of land had the highest rate of seasonal migration

Economic & Political Weekly

january 28, 2012 vol xlvii no 4

(29 per 1,000). With an increase in landholding, the migration rate decreased. The same did not hold true for urban areas. Education is a reflection Table 4: Temporary and Seasonal Migration Rate (Migrants Per Thousand) according to

of socio-economical develop-

Educational Status by Place of Residence ment and a prime determi-(Age-Group 15-64 Years), National Sample Survey, 2007-08

nant of the job a migrant

Educational Status Rural Urban Total

labourer is going to have.

Below primary 29.1 7.8 25.7

As Table 4 shows, temporary

Primary or middle level 27.9 6.2 22.1

migration rates decreased

Higher secondary level 16.8 3.9 11.1

with increasing levels of edu-Graduate or above 17.4 4.2 8.3

cation. We find that people Total 26.4 5.5 20.5

Source: Same as in Table 1.

who had less than primary education showed the highest propensity to migrate. This trend prevailed in rural as well as urban areas.

As seen in Table 5, temporary migration rates of different social groups varied considerably with it being very high among STs (45 per 1,000). The rate was 25 per 1,000 for SCs. In rural areas, STs had a higher rate

Table 5: Temporary and Seasonal Migration of temporary migration (49) Rate (Migrants Per Thousand) according to Social Group by Place of Residence (Age-Group

than SCs (30). Further, the

15-64 Years), National Sample Survey, 2007-08

results show that differen-Social Group Rural Urban Total

tials among social groups Scheduled tribes 49.0 6.5 45.2

were pronounced in rural Scheduled castes 29.9 6.7 24.8

areas but less remarkable in Other Backward Classes 23.9 6.5 19.5

Others 18.0 4.3 12.2

urban areas. This confirms

Total 26.4 5.5 20.5

our proposition that the

Source: Same as in Table 2.

poor and socially deprived in rural areas migrate more than others on a temporary basis. It could be said that poor and socially deprived classes are more mobile because of the distress-driven nature of temporary migration in India. Table 6: Temporary and Seasonal Migration

Rate (Migrants Per Thousand) according to

Table 6 shows temporary

Religion by Place of Residence (Age-Group and seasonal migration rates 15-64 Years), National Sample Survey, 2007-08

Religion Rural Urban Total

by religion. The migrations

Hindu 26.3 5.0 20.6

rate was highest among

Muslim 32.4 7.5 23.2

Muslims (23 per 1,000), fol-

Others 14.8 7.6 12.4 lowed by Hindus (20 per Total 26.4 5.5 20.5

1,000). The pattern was the Source: Same as in Table 1. same in rural as well as urban areas. Nevertheless, religious differences were more prominent in rural than urban areas.

4.3 Factors Associated with Temporary and Seasonal Migration

We fitted various logistic regression models to examine the effect of economic factors while controlling for other social-economic, demographic and geographical variables (Table 7, p 86). In the rural sample, a statistically significant negative relationship between MPCE tertiles and seasonal migration was observed, which implied that persons belonging to lower income groups were more likely to migrate temporarily (in model I).

Further, the likelihood of temporary migration declined with more land being owned by a household, which confirms previous research findings (Connell et al 1976). A plausible explanation may be that households with smaller landholdings try to diversify their activities through seasonal migration to supplement rural income in the agricultural lean season (Hugo 1985; Vanwey 2003).

Table 7: Results of Logistic Regression Analysis for Determinants of Temporary and Table 8: Results of Logistic Regression Analysis for Determinants of Temporary Seasonal Migration in Rural Areas (Age-Group 15-64 Years), National Sample Survey, 2007-08 and Seasonal Migration in Urban Areas (Age-Group 15-64 Years), National Sample

Survey, 2007-08

Covariates Rural (N=235682)

Model I Model II Covariates Urban (N=134922) Model III Model IV

MPCE tertile Low® 1.00 1.00

MPCE tertile

Low® 1.00 1.00

Medium 0.62*** 0.71***

High 0.36*** 0.53*** Medium 0.62*** 0.65***

Educational attainment High 0.52*** 0.58*** Below primary® 1.00 1.00 Educational attainment

Primary or middle 1.09*** 0.67*** Below primary® 1.00 1.00

Secondary or higher 0.79*** 0.47*** Primary or middle 0.88 0.58***

Land possession Secondary or higher 0.66*** 0.43*** Less than 1 hectare® 1.00 1.00

Land possession 1-4 hectares 0.70*** 0.77***

Less than 1 hectare 1.00 1.00 More than 4 hectares 0.48*** 0.54***

1-4 hectares 1.31 1.33 Social group

More than 4 hectares 2.91*** 3.70***

Scheduled tribes® 1.00 Social group

Scheduled castes 0.60***

Scheduled tribes® 1.00

Other Backward Classes 0.52*** Scheduled castes 1.71*

Others 0.47*** Other Backward Classes 1.70*

Religion Others 1.41

Hindu® 1.00 Muslim 1.21***

Religion Hindu® 1.00

Others 0.81* Size of the household

Muslim 1.30*

Less than 5® 1.00 Others 2.19***

5 or more 1.04 Size of the household SexLess than 5 1.00

Male® 1.005 or more 0.84* Female 0.13*** Sex

Marital status Male® 1.00 Single® 1.00

Female 0.17*** Currently married 1.57***

Marital statusState Single 1.00 Rajasthan® 1.00

Currently married 0.85* Andhra Pradesh 0.77**

State Assam 0.74*

Rajasthan 1.00 Bihar 2.00***

Andhra Pradesh 0.37*** Gujarat 2.09***

Arunachal Pradesh 3.54* Haryana 0.31***

Assam 3.83** Himachal Pradesh 0.50*

Bihar 1.60*

Karnataka 0.66***

Haryana 0.40*

Kerala 0.48*** Maharashtra 0.40***

Madhya Pradesh 1.30*** Nagaland 12.98***

Maharashtra 0.76*** Punjab 0.34**

Manipur 0.26* Uttar Pradesh 0.65*

Meghalaya 0.73* Tamil Nadu 1.43*

Nagaland 2.64** Orissa 0.59***

Jharkhand 0.20*

Uttarakhand 0.28*

Punjab 0.46***

Tamil Nadu 0.78** Agea 0.98*** *p<0.1, **p<0.05, ***p<0.001; ® reference category; a continuous variable; only statistically

Tripura 0.12***

significant states are shown in the table. West Bengal 1.28***

social group, religion, size of household, sex, marital status and

Chhattisgarh 0.61*** Jharkhand 1.38***state (model II). Nevertheless, those with less than primary edu Uttarakhand 0.35***cation were found to be highly mobile on a temporary basis com Agea 0.96*** pared to those with middle-level education or higher. It could be *p<0.1, **p<0.05, ***p<0.001; ® reference category; a continuous variable; only statistically

significant states are shown in the table. inferred that the chances of temporary migration in rural areas Source: Same as in Table 1.

went down with an increasing level of education. These results We also found a negative association between educational were consistent with earlier studies (Keshri and Bhagat 2010). It attainment and temporary migration. It showed that those with a can be seen that the logistic regression results endorsed the lower level of education had the highest propensity to migrate. bi-variate results relating to social group. STs were two times These relationships were not distorted even after controlling for more likely to migrate seasonally compared to SCs and other caste

86 january 28, 2012 vol xlvii no 4


groups. Muslims were significantly more likely to migrate than 5 Conclusions Hindus and others. The results also showed that males had higher Temporary and seasonal migration has long been an important odds of migrating than females. This could lead to the inference income diversification and risk-coping strategy in many agriculturethat males have higher chances of migrating temporarily when based economies in the developing world. In places where access to other factors are controlled for in rural areas. non-agricultural employment is limited, or climate (or technology)

The results suggested that people from Bihar and Gujarat were prevents continuous cultivation, seasonal migration is often the two times more likely to migrate temporarily or seasonally com-key to a household’s income during the agricultural lean season. pared to those from Rajasthan (which has a temporary migration It is not only an important form of labour mobility in a country rate equal to that of India). Rural inhabitants of Madhya Pradesh, with an increasing shift of the labour force from agriculture to West Bengal, Nagaland and Jharkhand were more likely to migrate industry and the tertiary sector (Keshri and Bhagat 2010), but also compared to the remaining states in the country. We found that critical to the livelihoods of socially deprived groups, especially those belonging to economically backward and low-growth tribal people and those from rural areas who lack of employment states had a higher likelihood of migrating seasonally. However, at their place of origin. This study presents regional patterns and, Gujarat, a high-growing state, was an exception to this. This was more importantly, the socio-economic determinants of tempoprobably because the state has had a history of seasonal migra-rary and seasonal migration in India more specifically. tion from its southern and south-eastern parts that consist of dry, Regional variations in temporary migration are noteworthy in hilly and tribal-dominated districts (Breman 1994; Krishna et al a country. Bihar, Jharkhand, Gujarat, Madhya Pradesh, West 2003; Jayaraman 1979). Bengal and Nagaland have a very high intensity of migration. All

The logistic regression results for the urban sample were more these states either have a high level of intra-state inequality or or less similar to that of rural areas with an exception pertaining a high proportion of STs and SCs. We observe stark rural-urban to landholding (in model III) (Table 8, p 86). It was observed that differentials in the intensity of temporary migration, which may those from households with more than 4 hectares of land had a be explained by differentials in levels of economic development in higher likelihood of migrating seasonally. After controlling for rural and urban areas and the resulting availability of employment. other socio-economic variables (in model IV), we find that in ur-Overall, temporary and seasonal migration declines with better ban areas as well the chances of temporary migration decreased economic and educational status. In rural areas, those with inwith increasing income. creasing incomes become less prone to migrate temporarily. Social

This result is in contrast to an earlier study pertaining to urban factors play a critical role in migration decisions. Those belongareas by Keshri and Bhagat (2010). Further, there was an impor-ing to STs have a higher chance of migrating seasonally than peotant deviation from the findings on rural areas in that SCs and ple in any other social group. This finding corresponds to earlier OBCs had higher chances of migrating temporarily in urban studies, which have documented that lower caste and tribal people areas. The same was true of lower educational levels, which trapped in poverty have a greater propensity to migrate seasonally meant urban inhabitants with less than middle-level education (Breman 1996; Deshingkar and Start 2003; Keshri and Bhagat had higher chances of migrating temporarily. The results 2004; Mosse et al 2005; Rao and Rana 1997; Rao 2005; Rogaly et suggested that urban people from Bihar, Tamil Nadu, Assam, al 2001). In short, the study concludes that temporary mobility is Arunachal Pradesh and Nagaland were more likely to migrate higher among the poorer sections of Indian society irrespective seasonally than from other states in the country. of the level of economic development of the states concerned.

Notes Development Report, Overseas Development Dayal, H and A K Karan (2003): Labour Migration Institute, London. from Jharkhand, Institute for Human Develop

1 Keshri and Bhagat (2010) point out that according

Brauw, A D (2007): “Seasonal Migration and Agricul- ment, New Delhi.

to earlier NSS migration-related data (1999-2000), ture in Vietnam”, ESA Working Paper No 07-04 Deshingkar, P (2006): “Internal Migration, Poverty a larger proportion (around 20%) of temporary (FAO: Agricultural Development Economics and Development in Asia”, Briefing Paper, Overmigrants belonged to the 0-14 age-group. Such

Division). seas Development Institute, London.

discrepancies are not found in the 64th round Breman, J (1978): “Seasonal Migration and Cooperative Deshingkar, P and S Akter (2009): “Migration and (2007-08).

Capitalism: Crushing of Cane and of Labour by Human Development in India”, Research Paper 2 We consider the working age-group (15-64) popu-

Sugar Factories of Bardoli”, Economic & Political 2009/13, Human Development Reports, UNDP. lation for the analysis because it is assumed that

Weekly, 13 (31), pp 1317-60. Deshingkar, P and J Farrington (2009): “A Framework temporary migration is employment-related. How

– (1994): Wage Hunters and Gatherers: Search for of Understanding Circular Migration” in P Deshever, for state-wise estimates, the whole sample Work in the Urban and Rural Economy of South

ingkar and J Farrington (ed.), Circular Migration

was considered. Gujarat (Delhi: Oxford University Press).and Multilocational Livelihood Strategies in Rural

3 Binary logistic regression models are used where – (1996): Footloose Labour: Working in India’s India (New York: Oxford University Press).

the dependent variable is in a binary form. Informal Economy (Cambridge: Cambridge Uni-Deshingkar, P and S Grimm (2005): “Internal Migration

versity Press). and Development: A Global Perspective”, Paper

Chandrasekhar, C P and J Ghosh (2007): “Dealing with for International Organisation for Migration, References Short-term Migration”, Business Line, 25 September. Geneva.

– (2010): “Migration for Work”, Business Line, 28 Deshingkar, P, R Khandelwal and J Farrington (2008): Banerjee, A and S Raju (2009): “Gendered Mobility:

December. “Support for Migrant Workers: The Missing Link Women Migrants and Work in Urban India”,

Chapman, M and R M Prothero (1983): “Themes on in India’s Development”, Natural Resource Per-

Economic & Political Weekly, 54 (28), pp 115-23.

Circulation in the Third World”, International spectives, 117, Overseas Development Institute, Bhagat, R B (2010): “Internal Migration in India: Are London.

Migration Review, 17 (4), pp 597-632. the Underprivileged Migrating More?”, Asia

Connell, J, B Dasgupta, R Laishley and M Lipton Deshingkar, P and D Start (2003): “Seasonal Migra-Pacific Population Journal, 25 (1), pp 27-45. (1976): Migration from Rural Areas: The Evidence tion for Livelihoods in India: Coping, Accumula-Bird, K and P Deshingkar (2009): “Circular Migration from Village Studies (Delhi: Oxford University tion and Exclusion”, Working Paper 220, Overseas

in India”, Policy Briefing No 4, Prepared for World Press). Development Institute, London.

Economic & Political Weekly

january 28, 2012 vol xlvii no 4 87

Gnanou, K M (2008): “Debt Bondage, Seasonal Migration and Alternative Issues: Lessons from Tamil Nadu (India)”, Autrepart, 30, pp 127-42.

Ha, W, J Yi and J Zhang (2009): “Inequality and Internal Migration in China: Evidence from Village Panel Data”, Research Paper, Human Development Reports, UNDP.

Haberfeld, Y, R K Menaria, B B Sahoo and R N Vyas (1999): “Seasonal Migration of Rural Labour in India”, Population Research and Policy Review, 18 (5), pp 471-87.

Hugo, G J (1982): “Circular Migration in Indonesia”, Population and Development Review, 8 (1), pp 59-83.

– (1985): “Circulation in West Java, Indonesia” in R M Prothero and M Chapman (ed.), Circulation in Third World Countries (Boston: Routledge and Kegan Paul).

Jayaraman, T K (1979): “Seasonal Migration of Tribal Labour: An Irrigation Project in Gujarat”, Economic & Political Weekly, 14 (41), pp 1727-32.

Keshri, K and R B Bhagat (2004): “Temporary and Seasonal Migration in India: The Magnitude, Pattern and Characteristics”, Malaysian Journal of Tropical Geography, 35(1-2): 17-30.

– (2010): “Temporary and Seasonal Migration in India”, Genus, 66 (3): 25-45.

Krishna, A, M Kapila, M Porwal and V Singh (2003): “Falling into Poverty in a High-Growth State: Escaping Poverty and Becoming Poor in Gujarat Villages”, Economic & Political Weekly, 38(49): 5171-79.

Kundu, A and N Sarangi (2007): “Migration, Employment Status and Poverty: An Analysis across Urban Centres”. Economic & Political Weekly, 42 (4): pp 299-306.

Lam, T Q, B John, R A Chamratrithirong and Y Sawangdee (2007): “Labour Migration in Kanchanaburi Demographic Surveillance System: Characteristics and Determinants”, Journal of Population and Social Studies, 16 (1), pp 117-44.

Mosse, D, S Gupta and V Shah (2005): “On the Margins in the City: Adivasi Seasonal Labour Migration in Western India”, Economic & Political Weekly, 40 (28): 3025-38.

Mukherji, S (2006): Migration and Urban Decay: Asian Experiences (New Delhi: Rawat Publication).

National Sample Survey Office (2001): “Migration in India”, Report No 470 (55/10/8), 1999-2000 (Ministry of Statistics and Programme Implementation, New Delhi).

– (2010): “Migration in India”, Report No 533 (64/10.2/2), 2007-2008 (Ministry of Statistics and Programme Implementation, New Delhi).

Pham, B N and P S Hill (2008): “The Role of Temporary Migration in Rural Household Economic Strategy in a Transitional Period for the Economy of Vietnam”, Asian Population Studies, 4 (1), pp 57-75.

Prothero, R M and M Chapman, ed. (1985): Circulation in Third World Countries (Boston: Routledge and Kegan Paul).

Rao, N (2005): “Power, Culture and Resources in Gendered Seasonal Migration from Santhal Parganas” in S Arya and A Roy (ed.), Poverty, Gender and Migration (New Delhi: Sage Publication).

Rao, N and K Rana (1997): “Women’s Labour and Migration: The Case of the Santhals”, Economic & Political Weekly, 32 (50), pp 3187-89.

– (2005): “Power, Culture and Resources in Gendered Seasonal Migration from Santhal Parganas” in S Arya and A Roy (ed.), Poverty, Gender and Migration (New Delhi: Sage).

Rogaly, B (1998): “Workers on the Move: Seasonal Migration and Changing Social Relations in Rural India”, Gender and Development, 6 (1), pp 21-29.

Rogaly, B, J Biswas, D Coppard, A Rafique, K Rana and A Sengupta (2001): “Seasonal Migration, Social Change and Migrants’ Rights: Lessons from West Bengal”, Economic & Political Weekly, 36 (49), pp 4547-59.

Singh, D P (2009): “Poverty and Migration: Does Moving Help?” in Amitabh Kundu (ed.), India: Urban Poverty Report 2009 (New Delhi: Oxford University Press).

Skeldon, R (2002): “Migration and Poverty”, Asia-Pacific Population Journal, 17 (4), pp 67-82.

Srivastava, R and S K Sasikumar (2003): “An Overview of Migration in India, Its Impacts and Key Issues”, paper presented at Regional Conference on Migration, Development and Pro-Poor Policy Choices in Asia, Dhaka.

Stark, O and D E Bloom (1985): “The New Economics of Labour Migration”, American Economic Review, 75 (2), pp 173-78.

Stark, O and D Levhari (1982): “On Migration and Risk in LDCs”, Economic and Cultural Change, 31, pp 191-96.

Vanwey, L K (2003): “Land Ownership as a Determinant of Temporary Migration in Nang Rong, Thailand”, European Journal of Population, 19 (2), pp 121-45.

Vijay, G (2005): “Migration, Vulnerability and Insecurity in New Industrial Labour Markets”, Economic & Political Weekly, 40 (22), pp 2304-12.

Yang, X (1992): “Temporary Migration and Its Frequency from Urban Households in China”, Asia Pacific Population Journal, 7 (1), pp 27-50.

Yang, X and F Guo (1999): “Differences in Determinants of Temporary Labour Migration in China: A Multilevel Analysis”, International Migration Review, 33 (4), pp 929-53.

Zelinsky, W (1971): “The Hypothesis of the Mobility Transition”, Geographical Review, 61 (2), pp 219-49.

Economic Reforms and Growth in India

Essays from Economic and Political Weekly Edited by


This volume investigates the nature of economic growth in India, its pace over time, its relationship to changes in the policy regime and the role of the external sector, and uses data to evaluate the policies that have implicitly underpinned the changes.

Presenting a range of approaches, views and conclusions, this collection comprises papers published in the Economic and Political Weekly between the late 1990s and 2008 that are marked by an empirical awareness necessary for an understanding of a growth history. The articles reflect a certain groundedness in their approach in that they privilege content/context over methodology.

This volume is an important addition to the literature on post-liberalisation economic growth in India. It will be useful to students and scholars of economics and management.

Contributors include Deepak Nayyar • Rakesh Mohan • Atul Kohli • Arvind Panagariya • Kunal Sen • Neeraj Hatekar • Jessica Seddon Wallack • Pulapre Balakrishnan

• Ravindra Dholakia • Ramesh Chand • R. Nagaraj • Montek Ahluwalia • Shashank Bhide • Amit Bhaduri • Pranab Bardhan

Readings on the Economy, Polity and Society

This is the first book in a new series jointly published by EPW and Orient Blackswan. The series is being published as part of a University Grants Commission project to promote teaching and research in the social sciences in India. The project (2010-12) is being jointly executed by the Tata Institute of Social Sciences, Mumbai, and the Economic and Political Weekly. The series is meant to introduce university students and research scholars to important research that has been published in EPW in specific areas. The readers draw on the EPW ’s archive of published articles.

Pp xiv + 454 ISBN 978-81-250-4271-6 2011 Rs 445

Orient Blackswan Pvt Ltd

Mumbai Chennai New Delhi Kolkata Bangalore Bhubaneshwar Ernakulam Guwahati Jaipur Lucknow Patna Chandigarh Hyderabad


january 28, 2012 vol xlvii no 4

To read the full text Login

Get instant access

New 3 Month Subscription
to Digital Archives at

₹826for India

$50for overseas users


(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top