TRAVEL TO WORK IN INDIAtripp.iitd.ac.in/assets/publication/WorkTravelReport.pdfTransportation Research and Injury Prevention Programme, the Indian Institute of Technology Delhi. This

TRAVEL TO WORK IN INDIA

Current Patterns and Future Concerns

Transportation research and Injury Prevention Programme (TRIPP)

Indian Institute of Technology Delhi

TEAM:

Geetam TiwariNishant

Acknowledgements:

“Travel to Work in India: Current Patterns and Future Concerns” has been prepared by the Transportation Research and Injury Prevention Programme, the Indian Institute of Technology Delhi. This report was partially supported by the TIGTHAT project, a UK Medical Research Council, Global Challenges (MR/P024408/1) award.

The research team benefited from the interaction with the participants at The Roundtable Discussion on Travel to Work in India: Current Trends and Future Concerns in IIT, Delhi in 2018. We are grateful to, among others, Prof. Dinesh Mohan, Pradeep Sachdeva, A.K.G. Menon, K.T. Ravindran, A.K. Dunu Roy, Shrikant Gupta, Amit Bhatt and Akshima Ghate. Thanks to Dr. James Woodcock, University of Cambridge for feedback on an earlier draft of this report. We would also like to thank Mr. Mahesh Gaur and other TRIPP staff members for providing assitance during the project.

© Transportation Research and Injury Prevention ProgrammeIndian Institute of Technology Delhi 2018-10-01

Suggested Citation:Tiwari, G. and Nishant (2018) Travel to Work in India: Current Patterns and Future Concerns. TRIPP-PR-18-01. Transport Research & Injury Prevention Programme, Indian Institute of Technology Delhi, New Delhi.

This Document is for public circulation. It may be reproduced in part with attribution to TRIPP, IIT Delhi.

TRAVEL TO WORK IN INDIA

Current Patterns and Future Concerns

Transportation research and Injury Prevention Programme (TRIPP)Indian Institute of Technology Delhi

Source: Unsplash

HIGHLIGHTS 1

Work Trips: India and it’s States 3Travel to work in India: Rural and Urban 4State-wise work trip characteristics 6

1.1 Trip length distribution for work trips at state level in India1.2 Modal share in work trips at state level 1.3 Variation in urban work travel characteristics 1.4 The “No-travel” workers

Work travel: Census data at district level 112.1 Work commute trends in different categories of urban districts 2.2 Vehicle ownership and travel patterns2.3 Mapping the travel distances

Gender-based differences in commute to work 223.1 Population Size, Density, and Gendered Work Travel 3.2 Gender and the ‘No-travel’

Suggestions 31Scope of the survey 32Data quality and reliability 32

References / Appendix 35

Appendix I - DefinitionsAppendix II - Classification of UA/TownsAppendix III - Population density distribution Appendix IV - Regions and Regional VariationAppendix V - Datasheets on statesAppendix VI - Census survey questions on travel to workplaceAppendix VII - Work location in NSS data (68th round)Appendix VIII - State-wise shares of establishment structure (6th Economic Survey)Appendix IX - Vehicle ownership in households

Contents

I

List of Figures

Figure 1.1: Trip lengths in rural and urban work trips 4Figure 1.2: Modal share in rural and urban work trips 5Figure 1.3: Modal share for work trips in urban India 5Figure 1.4: Distribution of trip length and mode of transport in urban India work trips 6Figure 1.5(a): Trip length distribution of work trips in seven regions of India 7Figure 1.5(b): Distribution of modes of transport to work trips in seven regions of India 7Figure 1.6(a): Trip length distribution across degrees of urbanization 8Figure 1.6(b): Distribution of modes of transport across degrees of urbanization 8Figure 1.7(a): Trip length distribution across levels of GSDPPC 9Figure 1.7(b): Distribution of modes of transport across levels of GSDPPC 9Figure 1.8(a): Trip length distribution across levels of highway density 9Figure 1.8(b): Distribution of modes of transport across levels of highway density 9Figure 2.1(a): Travel pattern in districts with population more than 8 million 13Figure 2.1(b): Percentage of households with particular type of vehicle in possession for use 13Figure 2.2(a): Travel pattern in districts with population between 4 and 8 million 14Figure 2.2(b): Percentage of households with particular type of vehicle in possession for use 14Figure 2.3: Travel pattern in districts with population between 2 and 4 million 15Figure 2.4: Travel pattern in districts with population between 1 and 2 million 15Figure 2.5(a): Travel pattern in districts with population between 5 and 10 Lakh 16Figure 2.5(b): Travel pattern in districts with population between 1 and 5 Lakh 17Figure 2.5(c): Travel pattern in districts with population less than 1 Lakh 17Figure 2.5(d): Travel pattern in districts with population less than 1 Lakh 18Figure 2.6: Scatterplots between ownership of bicycle and mode share of bicycle in various trip lengths 18Figure 2.7: Scatterplot between mode share of MTW and ownership of MTW 19Figure 2.8: Scatterplot between mode share and ownership of car 19Figure 2.9: Percentage of very short trips (< 1 km) among all the work trips 20Figure 2.10: Percentage of short trips (2-5 km) among all the work trips 20Figure 2.11: Percentage of medium-long trips (6-20 km) among all the work trips 21Figure 2.12: Trip length distribution of work trips across the districts of India 21Figure 3.1: Gender-wise work trip pattern in Indian ‘mega-districts’ 24Figure 3.2: Trip length distribution for women 24Figure 3.3: Gender-wise work trip pattern in other Indian districts 25Figure 3.4: Gender difference with respect to modal shares of different modes of travel 26Figure 3.5: Comparing the ‘no travel’ among male and female workers 27Figure 3.6(a): Distance-wise percentages of the male and female workers who walk to work 28Figure 3.6(b): Percentage of male or female workers who ride a bicycle to workplace 28Figure 3.7: Percentage of no-travel workers among the women minus that among the men 29Figure 3.8: Spatial distribution of no-travel workers in urban India 30Figure 3.9: Spatial variation in Sopher’s Gender Disparity Index (DIS) of ‘no-travel’ 30Figure A-1: Bivariate distribution of population and population density (aggregated at district level) 39Figure A-2: Histogram of population density in India (densities aggregated at district level) 40Figure A-3: Poverty levels (percentage of population below poverty line) vs. Area of State 41Figure A-4: Difference between ‘percentage of no-travel recorded in the Census’ and ‘percentage of workplace inside household recorded in the Economic Census’ 50

II

List of Tables

Table 1.1: Urbanization criteria used for classification of states 8Table 2.1: Classification of districts based on demographic criteria of size and density of population 12Table 3.1 Correlation analysis of no-travel and other variables at the district level 29Table A-1: Number of districts belonging to each population category 39Table A-2: List of regions- grouping the neighboring states 40Table A-3: Trip length distribution of work trips in different regions of India 40Table A-4: Mode share of work trips in different regions of India 41Table A-5: States and UTs in descending order of urbanization 42Table A-6: Highway road density in states (descending order) 43Table A-7: Urban poverty in states (ascending order) 43Table A-8: Per capita GSDP of Indian States/UTs (at constant prices)* 44Table A-9: State-wise percentage of workers in different trip length categories 45Table A-10: State-wise percentage of work trips in different travel mode categories 46Table A-11: Location of workplace categories used in NSS 68th round] 47Table A-12: Classification of workers in different occupational divisions based on NCO 2005 48Table A-13: State-wise distribution of establishment structure* 48Table A-14: Percentage of households possessing particular type of vehicle 51

How to workers travel for work in India?

1

• Census 2011 is the first Indian census to record any sort of travel information. It will provide a baseline to assess the change in pattern in the coming decades. Despite many limitations, district-level information on work trips is precious since it is the only published nationally-representative information on the work commutes of all Indian workers.

• A plurality of Indian workers travel on foot or by bicycle (58.1% and 48.9% workers in rural and urban areas, respectively). The share of motorized two-wheelers in urban districts is nearly 20% while the modal share of the car (or a jeep/van) is less than 5%. Users of para-transit have a minimal share (less than 5%). With more than 20% share, use of public transportation modes (bus and train) for travel to work is significant in urban as well as rural districts.

• The proportion of workers who travel a distance of less than 5 kilometer is almost 70% for both rural and urban districts. Only a little above 10% workers in rural as well as urban areas have work location farther than 10 kilometers.

• The proportion of trips involving longer distances (more than 10 km) is very low across all the states. The proportion of trips involving more than 10 km is highest in NCT of Delhi (25%) and Goa (22.8%).

• 24.5% workers in urban areas and 38.8% workers in rural areas do not travel at all to get to their workplaces. Nearly one-fifth of male workers and one-third women workers in urban India do not leave home for their work.

• Northern states have the highest ratio of no-travel workers but the overall trip length distribution doesn’t vary much across the geographical zones. The states in South India have the highest share of bus travel (27%). Walking has a highest modal share in the North-eastern states (47%).

• No-travel is higher in the states with lower per capita GSDP. Also, the richer the state is higher is the share of longer trips and higher is the use of private automobiles. The modal share of bicycles is lowest (but never less than 10%) in the richest states. The modal share of the car is significant only in the richest states where it is just above the 10% mark.

• Bicycling has a higher modal share in the states with higher levels of urban poverty. • Bicycling is more common in mid-size districts with moderate to high densities and so is the

case with motorbikes. Interestingly, both of these find very low usage among the women and these are the two most gender unequal modes for work travel.

• In the context of not traveling to work, there is remarkable variation in gender difference across the different parts of the country. There are 26 out of 640 districts where the percentage of workers not traveling among men is at least 5% higher than that among women. Only one of these districts (Bijapur, Maharashtra) is located in non-hilly terrain.

• Higher density and lower size are associated with greater proportion of short trips and walking.

HIGHLIGHTS

Getting answers from the Census 2011

2

• No-travel is highest in the largely rural states. The proportion of longer than 5 km trips is highest in the most urbanized states. The share of walking is lowest (yet more than 20%) in the most urbanized states. Use of informal public transport (IPT) is very low in rural as well as the most urbanized states.

• Use of public transport is high in districts with bigger population size as well as in those with low density. Traveling among female workers is contrastingly higher in low-density districts and the increased proportion of female travelers are using public transport to reach to the workplace.

• The car has the lowest modal share (not more than 8% in any state, except in Chandigarh (14.6%), Delhi (12.8%), and some hill states) among all the modes of transport. The bicycle has a higher modal share (more than 10%) in all states, except the hill states, Goa (3%), Kerala (5%), and Karnataka (8%).

• Although car use is highly income inequitable and a minority mode, in certain district categories, the proportion of women workers who drive themselves in the car to work is greater than the same proportion of men.

• Wherever the modal share of two-wheelers (bicycle and motorbike) is higher, there emerges difference in the trip length distribution of work trip by male and female workers. Either these districts have different nature of job/employment structure, or there is a lack of accessibility to well-paid employment or both. In any case, the significance of gender-transport-employment linkage can hardly be overstated

• Districts of different population size and density show different work travel characteristics. The metropolitan districts with big size and highly dense urban space have distinct travel patterns that are different from the patterns in districts with moderate size and density, and the districts with low size and density also have characteristically different travel patterns.

• The percentage of female workers who walk to work is higher than the percentage of men who walk to work in all of the 24 district categories, irrespective of trip length.

• Districts in Uttar Pradesh, Bihar, and Rajasthan have the highest gender difference in the rate of no-travel. An extremely low proportion of women workers travel to work in these states.

• Census should not only broaden the scope of the questions to include the non-work trips undertaken by each household member on a regular basis but it also needs to improve its questions so that they do not have ambiguity. Distance is difficult to measure with the tools made available to a surveyor and thus reported distances are perhaps only the best guesses of the actual distance involved. Complimenting it with information about travel time could enhance the data quality.


3

WORK TRIPS: INDIA AND IT’S STATES

Source: Unsplash


4

Census 2011 is the first census to obtain basic information about how Indian working population travels to their workplaces. So even if there is no data to understand the change in work travel patterns in India over the years, this can be a baseline to assess the trend in the coming decades. According to the Census, ‘other workers’ include all the people engaged in some economic activity, except cultivation, agricultural labor, and household industry.

Travel to work in India: Rural and Urban

Census data shows that 24.5% workers in urban areas and 38.8% workers in rural areas do not travel at all in getting to their place of employment. In rural India, one-third of male workers and more than half of female workers do not travel to work whereas nearly one-fifth of male workers and one- third women workers in urban India do not leave home for their work. Interestingly, the proportion of workers who do not travel to a distance of more than 5 kilometer is almost 70% in the case of both rural and urban workers. Only a little above 10% workers in rural as well as urban areas have work locations farther than 10 kilometers.

The larger portion of Indian workers travel on foot or by bicycle (58.1% and 48.9% workers in rural and urban areas, respectively). In urban areas, while little more than 20% workers travel to work by motorized two-wheelers, the share of those traveling to work by car (or a jeep/van) is less than 5%. According to the data, there is no gender divide in the proportion of workers traveling by car. Users of para-transit are minimal (less than 5%) among the workers on their trip to the job. Use of public transportation modes (bus and train) for

Figure 1.2 shows the modal share of work trips in rural as well as urban areas of India. Work trip mode shares are calculated as the proportion of workers choosing

to travel by a particular mode among the workers who actually travel to work (i.e. workers who do not make a

work trip are excluded)

Total

Rural

Urban

Figure 1.1: Trip lengths in rural and urban work trips


5

Figure 1.3: Modal share for work trips in urban India

Male Female All

travel to work is also significant for urban as well as rural India. More than 20% workers in both sectors have reported traveling by either of the two modes. Female workers in urban areas have an even higher share of riders of public transport. Though women workers have higher mode share of pedestrians (46.5% in rural and 67.6% in urban) than their male counterparts (28.8% in rural and 28.3% in urban) only 4.5% women workers ride bicycle to reach their job locations as compared to 20.2% men in urban areas. Chapter 3 provides a detailed discussion on the gender dimension of work travel.

Aggregate trip lengths and modal shares seem not to vary much across the rural and urban sectors. However, this report limits its scope to urban work trips. Hence in the later sections, only the urban aspect of work trips are discussed in more detail.

Total Rural Urban

Figure 1.2: Modal share in rural and urban work trips


6

State-wise work trip characteristics

In this section, trip length and mode share have been described on the smaller geographical scale of states and across several socioeconomic indicators. Interesting patterns of similarity, as well as variation, emerge. A brief summary of trip lengths and mode shares for urban areas in each of the 28 states and the 7 UTs are listed in Appendix A9 and A10 respectively.

1.1 Trip length distribution for work trips at State level

Firstly, the relative number of ‘no-travel’ workers varies considerably across the states- 10% in Daman & Diu and Puducherry while it is close to 50% in several states including Uttar Pradesh and Bihar. Census data on the occupational distribution of workers based on the 9 major divisions specified by National Classification of Occupation-2005 (Appendix VII) reveals that the largest number of workers are working in shops/market sales (15%) and craft & allied trades (21%). From the data, it can be assumed that a significantly high number of workers in the ‘no-travel’ category must be engaged in activities under these two occupational divisions (State-level data was not available at the time of drafting this report, so we could not relate the type of occupation with travel characteristics). Also, 55% workers in shop/market sales are not employees (i.e. they are either employers or single worker or family worker). This indirectly refers to the extent of the informal economy in urban India.

Let us define - trips within a kilometer as ‘short trips’, trips in the range of 2-10 km as ‘medium-long trips’, trips involving 10-30 km as ‘long trips’, and those to distances more than 30 km as ‘very long trips’. Except for smaller states/UTs, all the states have nearly the same proportion of trips involving short distances. The case of Chandigarh and Puducherry are

Figure 1.4: Distribution of trip length and mode of transport in urban India work trips


7

notable for a high share of medium length trips (60% and 55% respectively). Again, apart from the hill states, UP, Bihar and Andhra Pradesh have a lower proportion of medium length trips as compared to the other states. Long trips are significant mainly in the case of Delhi (more than 20%). Oddly enough the share of long trips is high in Goa and Puducherry (16% and 14% respectively). Very long trips of more than 30 km are in all the states and only relatively higher in Haryana (8%), Maharashtra, Tamil Nadu and West Bengal (7% each) - states where the four largest metropolitan cities are situated.

1.2 Modal share in work trips at State level

Among the workers who do travel to work, the share of pedestrians is high in most states, the only exceptions being Chandigarh and Puducherry which have less than 20% share of pedestrians. This makes sense since these are also the regions with a very low share of short trips (refer to the previous section). The highest share of pedestrian work trips is in the hill states, Bihar and MP (40% in both), and Daman & Diu (more than 60%). The share of bicycle trips shows larger variations across the states. States such as the hill states, Karnataka, Kerala, Delhi, and Maharashtra have even less than 10% share of bicycles among modes opted for travel to work. Comparing the combined NMT users across all the states, none of them has less than one-third share of NMT users. In addition to some smaller states/UTs, Uttar Pradesh (63%), Bihar (70%), Chhattisgarh and Odisha (67% each) have the highest modal share of NMT travelers. Motorized two-wheelers are a major portion of modes used for work trips in Gujarat, Puducherry (more than 30%), Punjab, Chandigarh, Chhattisgarh (25%), Rajasthan and Madhya Pradesh (22%) while its modal share is less than 10% in West Bengal and the hill states. Travel by car is low except in the case of Sikkim (21%), Chandigarh (15%), NCT of Delhi and Meghalaya (13%). The total modal share of personal motor vehicles for work trips is highest in Chandigarh, Puducherry, Goa (nearly 40% in each), and Gujarat (35%) while it is lowest in West Bengal (8%), Bihar (13%) and some of the hill states.

Major variations across the states are also perceptible in the use of public transport for traveling to work. Use of buses was most prominent in J&K (nearly 40%), Kerala, Himachal Pradesh, Goa, Tamil Nadu, Karnataka, NCT of Delhi (all having more than 25% modal share of buses). Modal share of the train is minuscule in all states except Maharashtra (15%) and West Bengal (13%). The total share of public transport is highest in J&K and Kerala (40%) while it is low in most states, especially in the hill states (except HP), Chhattisgarh, Chandigarh, Jharkhand, Gujarat, Odisha (less than 10%), MP, Punjab, UP and Bihar (10-15%).

1.3 Variation in Urban work travel characteristics

Figure 1.5(a): Trip length distribution of work trips in seven regions of India

Figure 1.5(b): Distribution of modes of transport to work trips in seven regions of India


8

states. Trains are a significant mode of travel to work only in the east and west.

Urbanization & No-Travel:Proportion of workers not traveling to work varies by level of urbanization in the states1 (Figure 1.6a). Nearly one-third of the workers in the least urbanized states do not commute while the share of such workers in most urbanized states is around 15%. The relation between urbanization and very short trips of less than a km is not clear though there is much variation. However, the proportion of longer trips (6-20 km) is clearly higher in ‘urban’ states as compared to the less urbanized states.

Looking at the variations in mode share (Figure 1.6b), most diverse is walking to work. Urban states have a decent share of walking (21%) but walking is considerably higher in semi-urban and other relatively rural states (30-45%). The share of the bicycle as a mode to work is between 10 and 25% but with no clear relationship with the degree of urbanization. The share of motorized two-wheelers also has the same range but is distinctly higher in the urban than rural states. Except in the urban states, where the share of the car is 14%, all other categories of states have a share of the car well below 10%. IPT modes, as has been noted previously in the report, have a very low (less than 5%) share irrespective of region and classification. The share of buses 1 See Appendix V for urbanization in different states

Figure 1.6(a): Trip length distribution across degrees of urbanization

Share of urban Population (%)

State of Urbanization

> 80 Urban

65-80 Mostly Urban

35-65 Moderately Urban

20-35 Mostly Rural

< 20 Rural

Table 1.1: Urbanization criteria used for classification of states

Zonal variation across the groups of geographically neighbor states:India can be loosely divided into seven regions by grouping adjacent states. See appendix for a list of states constituting each of the seven geographical neighborhood regions. Though there is a perceptible variation of the proportion of ‘no-travel’ workers, the trend of trip length distribution doesn’t seem to vary much across the regions. Interestingly, the proportion of workers who do not travel to work is not less than 20% in any region while being highest in North India.

Modal share of pedestrians is high for each region but the modal share of bicyclist varies considerably- it is high in the eastern region (30%) and low in the south, west and island regions (below 15%). Use of MTW also varies across the regions- highest in the west and reasonably high in other regions, except in east and northeast. Use of cars and para-transit modes is invariably low. Buses are most prominent in the South and island regions where these have almost one-fourth modal share while their modal share is less than 10% in the central Indian

Figure 1.6(b): Distribution of modes of transport across degrees of urbanization


9

Economic status- per capita GSDP:States have been classified based on quintiles. (Appendix V for more information on economic status of each state)Proportion of small trips(< 5kms) is higher than any other category of trips in all states, irrespective of the state economic standards. States with the highest economic output show a notably higher proportion of longer trips (6-20 km) as compared to other states. Percentage of trips longer than 20 km is considerably small in all economies.Walking is the predominant means of traveling to work in all kinds of economies with the proportion of walking among all modes being only a little higher in states of lower per capita economic output. In contrast to this, mode share of the bicycle is much higher in states of low per capita output (ranges from nearly 10% in highest GSDPPC states to 30% in lowest ones). Interestingly, mode share of buses is around 20% in all economies, except in those with

Figure 1.7(a): Trip length distribution across levels of GSDPPC

Figure 1.7(b): Distribution of modes of transport across levels of GSDPPC

is similar to bicycles and two-wheelers.

the lowest GSDPPC. Among the privately owned motor vehicles, the share of motorized two-wheelers doesn’t vary much with the level of economic output (15-20%) and the share of cars is notable only in states with the highest GSDPPC (and there too it is considerably low at 12%). Highway density:A greater density of highways does not correspond to longer work trips. Data rather shows that the proportion of smaller trips is more in states with greater highway density than those with a lower density. One notable insight is that the share of the bicycle is higher in states with a lower highway density whereas the mode share of buses is higher in those with denser highway network.

Figure 1.8(a): Trip length distribution across levels of highway density

Figure 1.8(b): Distribution of modes of transport across levels of highway density


10

1.4 The “No-travel” workers There are workers who have reported no regular travel relating to work (within the sample those who chose to report). These workers should be understood as those working in petty shops (which may be owned by them and located within the premises or in very close vicinity of their place of residence), working from home or self-employed in any other type of activity. We have earlier discussed how their proportion varies across the states and regions and the various socioeconomic attributes. In this section we discuss the significance of their size and if it provides any insight into the state of employment and the economy. One-fourth of the urban workers, including almost one-third of female workers and one-fifth of the male workers, have reported no-travel for work do not travel at all in relation to their job. How to make sense of these numbers? Are these figures high or low? Do these numbers tell us something interesting about the kind of urban economy India has? Are they worrying? The Census of India could have helped by providing disaggregate household-level profiles of workers and their families. The only choice is to look at the other limited source of data available- National Sample Survey (NSS).

The NSS (68th round) asked the workers a question about the location of the workplace- whether it is in his/her own dwelling unit, or attached to his/her own dwelling, or in an employer’s dwelling unit, and so on (See Appendix VII). The results showed about 15% workers either work in their own dwelling, or live in close proximity to work or do not have any fixed place of work. There is a gap of almost 10% between estimates of the two datasets. The same NSS also estimates the percentage of the self-employed working population in India (rural as well as urban) at nearly 50%. This is because NSSO’s definition of the ‘self-employed’ is independent of the location of the workplace. It includes all the own-account workers, employers, and helpers under the category of ‘self-employed’.

The Economic Survey is another national survey which can be of help in knowing about on the economic activities that people are engaged in. The 6th Economic Survey which was conducted in 2012-13 reported that 25.0% establishments in urban India were ‘inside household’ (Ministry of Statistics and Programme Implementation, 2016). This number is pretty close to 24.9% ‘no-travel’ in urban India reported by the Census 2011. Pearson’s correlation test run with ‘inside household’ and ‘percent no-travel’ at state-level produced positive coefficient of 0.40 which was significant at p=0.01. Though, across the states, the Census figures seem to be consistently inflated as compared to those of the Economic Census (see Appendix VII). Appendix VIII lists the state-wise percentage of establishments ‘inside household’ and ‘outside household without fixed structure’.

The Socio-Economic and Caste Census of India (SECC), another national survey, reports following as the distribution of (probably ‘no-travel’) workers at all India level- i) home-based workers- 6.8%(ii) shopkeepers/sales (‘no-travel’ is not necessary)– 7%(iii) some fraction of the 52% in ‘other work’ (about which no further details are available).

A major limitation of the SECC data on worker classification is that quite a large proportion of the working population falls under the category of ‘other worker’.


11

WORK TRAVEL: CENSUS DATA AT DISTRICT LEVEL

Source: Unsplash


12

In this section, descriptive analysis of census data at the district level is presented. The discussion here is based on district-level information and not the city-level information since the Census has neither provided the commute information at town level nor have they made disaggregate data available. It is acknowledged here that trip patterns are usually discussed at the city level and the limitation due to the scale mismatch between district and city is difficult to overcome. Such inconsistency in the spatial unit across different studies also adds to the difficulty while drawing comparisons and making comparative inferences. However, despite such limitations, district-level information on work trips is precious since it is the only information on the work commutes of all Indian workers. Census also collects information on the availability of each type of vehicle (bicycle, two-wheeler, car/ jeep/ van) in households. Information on vehicle availability is aggregated at the district level. There are 640 urban districts in India as per Census 2011. Census definition of urban district is given in Appendix I. This study classifies districts in terms of two demographic parameters- population and population density (see Appendix III for criteria limits and classification). The objective of such classification was to figure out patterns, if any, of trip length and mode share across the urban demographic contexts.

POPULATION POPULATION DENSITY NUMBER OF DISTRICTS

> 80 Lakh Very High 2

High 1

40 - 80 Lakh Very High 2

High 4


High 11

Medium 12

Low 2


High 17

Medium 34

Low 5


High 16

Medium 72

Low 14


High 48

Medium 162

Low 72

< 1 Lakh Very High 2

High 5

Medium 52

Low 71

Going by such detailed classification, there emerge 24 categories of districts! In the following section, travel pattern in each category is discussed.

Table 2.1: Classification of districts based on demographic criteria of size and density of population


13

2.1 Work commute trends in different categories of urban districts

This section presents a comparative discussion of travel patterns in districts classified as above specification. Many cities such as Mumbai, Kolkata, and Delhi consist of multiple districts. In this chapter, we are restricting the discussion to urban areas administratively defined as districts. Travel patterns in major cities will be discussed in the next chapter. We proceed in decreasing order of population size i.e. from the most populated to the least populated districts. The working population represents only those people who chose to take part in the study.

Districts with population more than 8 million:These districts are Mumbai Suburban, Bangalore (both in very high-density bracket) and Thane (high population density). percentages of workers out of the total population in these districts are respectively 38.2, 41.7 and 37.1. Figure 2.1(a) shows that the predominant modes of travel to work in these districts are NMT and public transport*. Less than 10%of work trips involves a car. Density doesn’t seem to make much difference here except that share of very long trips involving distances more than 20 km is quite high in the district with lower density. These very long trips are almost all undertaken by public transport. Figure 2.1(b) presents the vehicle available for use in households. What is worth noting here is that despite lower ownership of two-wheelers, the share of these vehicles in travel to work is higher in denser districts.

Figure 2.1(a): Travel pattern in districts with population more than 8 million

Figure 2.1(b): Percentage of households with particular type of vehicle in possession for use

*Public transport here doesn’t necessarily mean that transport service is provided by government agencies, or is public funded. Public transport as used here refers to all buses and trains including those run by private operators.

Working Population: 7038604 Working Population: 3051134


14

Districts with population of 4 to 8 million:

These districts are Ahmadabad, Pune, North Twenty Four Parganas, Surat (all with high population density), Chennai and Kolkata (both with very high population density) in decreasing order of population size. Proportionate size of working population in all these districts ranges from 33 to 40%. Here also NMT is the predominant mode of travel, more so for small trips. Motorized two-wheelers hold significant mode share in these districts as opposed to the case in districts with a very high population. MTW vehicles are mostly used for medium to long trips (2-20 km). Public transport has greater mode share in districts of very high density. Very long trips (more than 20 km) are small in proportion. Figure 2.2(b) presents the vehicle available for use in households. What is again interesting here is that despite higher ownership of two-wheelers, the share of these vehicles in travel to work is lower in denser districts (compare this with trend discussed in the previous section). The relationship between vehicle ownership and work travel has been described in more detail in the subsequent section (refer section 2.2).

Figure 2.2(a): Travel pattern in districts with population between 4 and 8 million

Population- 4-8 millionPopulation density > 10000 person/sq. km

Population- 4-8 millionPopulation density 5-10000 person/sq. km

Figure 2.2(b): ercentage of households with particular type of vehicle in possession for use

Districts with population 2-4 million:

A total of 31 districts have a population size of 2-4 million. If compared to the districts previously discussed, these districts have distinctively higher mode share of motorized two-wheelers and slightly lower mode shares of NMT. Across the varying levels of population density, districts of this population size have unchanging trip length distribution with most work trips involving medium-long trips (2-20 km). Very long trips are small in share. It is also noticeable that share of very short trips (less than a km) is consistent at around 20% across the density classification.



15




Population- 2-4 millionPopulation density < 2000 person/sq. km

Figure 2.3: Travel pattern in districts with population between 2 and 4 million

Districts with population 1-2 million:

Figure 2.4: Travel pattern in districts with population between 1 and 2 million




Working Population: 13154411Working Population: 2236329




Population- 1-2 millionPopulation density < 2000 person/sq. km


16

In this category of districts, an even higher fractions of trips are in the range of 2-5 km and the share of very long trips involving distance, more than 20 km, is very low. Trip patterns do not seem to vary with levels of population density. Trip length and mode share- both are almost identical across the density levels. Most of the very short trips (less than 1 km) and medium-short trips (2-5 km) are NMT trips. Among long trips of 6-20 km, NMT and motorized two-wheelers have a nearly equal mode shares. Mode share of buses is relatively lower than the districts with a population of more than 2 million which has been discussed in the earlier sections.

Districts with population 5-10 Lakh:A Greater share of medium-short trips (2-5 km) are a phenomenon in districts of the intermediate population size of 5-10 Lakh and their share is higher in districts of lower density. Most trips involve NMT but longer trips have increasingly more share of motorized two-wheelers. Use of buses is noticeable mainly in case of longer trips of more than 6 km.

Figure 2.5(a): Travel pattern in districts with population between 5 and 10 Lakh

Districts with population 1-5 Lakh:In contrast with districts of high population, the work trips in districts with a lower population size of 1-5 Lakh across the density levels are predominantly small in length (less than 5 km) and the NMT is mode most commonly used. Very small trips are highest (45%) in districts of very high density and this share decreases with lower density levels (but not receding below 30%). The opposite is the trend for 2-5 km long trips, a fraction of which rises with decreasing density levels; nearly 40% trips in low-density districts involve a distance of 2-5 km. Very long trips are again less than 10% of all trips across density levels but the majority of these trips are undertaken by buses while buses have a negligible mode share in other trips. This might be an indicator of workers living in suburban towns traveling to workplaces situated in the city by



Population- 5-10 LakhPopulation density > 10000 person/sq. km

Population- 5-10 LakhPopulation density 5-10000 person/sq. km

Population- 5-10 LakhPopulation density 2-5000 person/sq. km

Population- 5-10 LakhPopulation density < 2000 person/sq. km


17

accessing the inter-city bus services which otherwise are not available for travel inside the city.

Figure 2.5(b): Travel pattern in districts with population between 1 and 5 Lakh

Districts with population less than 1 LakhThe smallest districts with a population size of less than 1 Lakh have an incredibly high share of small and medium-small work trips and an even higher share of those undertaken without any motorized vehicle. It is interesting to notice not the identical pattern in the last three of the four charts in figure 2.5. Except for the districts with very high density having 55% trips in 0-1 km range, the other districts have identical trip length distribution and identical mode shares. What this means is that density might not have an important role in shaping trip patterns in these small districts (or cities).

Figure 2.5(c): Travel pattern in districts with population less than 1 Lakh




Population- 1- 5 LakhPopulation density > 10000 person/sq. km

Population- 1 -5 LakhPopulation density 5-10000 person/sq. km

Population- 1 -5 LakhPopulation density 2-5000 person/sq. km

Population- 1 -5 LakhPopulation density < 2000 person/sq. km

Population- less than 1 LakhPopulation density > 10000 person/sq. km

Population- less than 1 LakhPopulation density 5-10000 person/sq. km


18

Figure 2.5(d): Travel pattern in districts with population less than 1 Lakh

2.2 Vehicle ownership and travel patterns

Appendix IX lists the vehicle ownership as per the classification of districts used in this report. However, it is important to underline here again that these numbers are not necessarily the percentage of households actually holding the ownership of particular vehicles. Census records their data in terms of availability for use rather than ownership. So these numbers are a percentage of households where at least one vehicle of a particular kind is available with the household for use by its members. This part of the report explores the possible relationship between household-level vehicle ownership aggregated for each district and the work trip patterns. Vehicle categories used by the census are bicycle, two-wheeler (motorbike, scooter, moped etc.) and four-wheeler (car, jeep, van etc.).

Bicycle ownership and mode share Figure 2.6 shows the scatter plot between mode share of bicycles and bicycle ownership aggregated at the district level. Clearly, the relationship is concave-up parabolic suggesting that districts with a greater degree of bicycle ownership have increasingly higher mode share of bicycles in work trips. Three charts following this figure show break-ups of bicycle’s mode share by trip length. The difference in slopes of the first and the last two of these graphs needs attention. Though the mode share of bicycle increases for very small trips (0-1 km) too, the increment is quite flat as compared to the trend in medium to long trips. What this means is that greater proportion of those owning a bicycle do not play as instrumental a role in riding a bicycle to nearby work locations as they do to comparatively far away locations. Very small trips most usually tend to be walking trips.


Population- less than 1 LakhPopulation density 2-5000 person/sq. km

Population- less than 1 LakhPopulation density < 2000 person/sq. km

Figure 2.6: Scatterplots between ownership of bicycle and mode share of bicycle in various trip lengths


19

Figure 2.6: Scatterplots between ownership of bicycle and mode share of bicycle in various trip lengths

Ownership of motorized two-wheeler and travel to work

Having possession of a two-wheeler shows linear and positive relationship with mode share of two-wheeler in work trips. Approximately, for each 2% increase in a number of a household having two-wheelers, the mode share of MTW increases by unit%.

Figure 2.7: Scatterplot between mode share of MTW and ownership of MTW

Ownership of car and travel to work

Figure 2.8 shows the scatterplot between mode share and ownership of cars aggregated at the district level. Though less pronounced the relationship is positive meaning that it increases in the proportion of households having a car also increases the mode share of cars in overall work trips. Mode share of cars is higher than 15% in only those districts that have a relatively high car ownership. Also, on an average, 5% increase in households having a car is related to 1% increase in the mode share of cars. Therefore, car ownership does not necessarily get translated into use for car of work travel. Figure 2.8: Scatterplot between mode share and

ownership of car


20

2.3 Mapping the travel distances

In this section, trip length distribution in all the 640 districts (except Kinnaur, Lahul & Spiti, and Nicobars i.e. the districts where the number of workers reported is nil) is presented in form of choropleth maps. The maps shown here include the region of India where the Census has recorded at least one worker. The intention is not to misrepresent the political boundary of India or the administrative boundary of its districts. The proportion of very short trips (trip length less than 1 km) among all the work trips is high in the hilly districts and it is also high among the female workers in central and eastern states of India (see figure 2.9).

The proportion of short trips (trip length between 2 and 5 km) among all the work trips is highest among all the trip length categories across almost all the regions of the country (except the hill states) and this holds true for both men and women (refer to figure 2.10). The proportion of short trips is especially much higher in some regions of the north-west and the central India.

The proportion of medium-long trips (trip length between 6 and 20 km) is low in nearly every district (refer to figure 2.10). In most parts of the country, the proportion of the short trips is the highest among all trip length categories (refer to figure 2.12).

Figure 2.9: Percentage of very short trips (< 1 km) among all the work trips

Figure 2.10: Percentage of short trips (2-5 km) among all the work trips

All Male Female

All Male Female


21

All Male Female

Figure 2.11: Percentage of medium-long trips (6-20 km) among all the work trips

No Travel Very Short Trips (<1km)

Short Trips (2-5km)

Medium Trips (5-10km) Long Trips (10-20km) Very Long Trip (>20km)

Figure 2.12: Trip length distribution of work trips across the districts of India


22

GENDER-BASED DIFFERENCES IN COMMUTE TO WORK

Source: Unsplash


23

At 31% according to census-2011, India has one of the lowest female workforce participation rates in the developing world. Gender gap in participation rate averages around 60 percentage points in urban areas.The previous chapter presented a broad pattern of work trips at the district level. Now, this chapter aims to provide a comparison between the travel characteristics of male and female workers in India. Starting with the description of gender differences in trip length and modal share, major gender disparities and concerned issues are discussed.

Figures 3.1 and 3.3 show the variation in work travel across the categories of districts and the gender of the workers. The values on the y-axis are percentage of workers according to different trip length and modal categories. Unlike the percentages in charts shown in the previous chapter, these numbers are not the percentages of trips. Rather, these numbers are the percentage of all the workers including the ‘no-travel’ workers. Though the section 3.3 in this chapter provides a separate discussion on the ‘no-travel’ workers, the height of bars in the charts of figures 3.1 and 3.3 is an indirect indicator of the fraction of such workers.

Figure 3.1 is a presentation of work travel in mega districts (large size with higher densities). For each of the four categories of such mega districts, major differences between travel lengths of men and women are in medium and long trip lengths. The proportion of women as compared to the men who travel to distances longer than 5 km is lower. The two most gender unequal modes are bicycle and motorbike which have comparatively low ridership among women. The public transport (mainly consisting bus) and the car appear to be the most gender equal modes while more workers among the women walk to the workplace as compared to those among the men. Modal share of the bicycle is higher for districts with a population of 4-8 million although the gender divide persists. Similar is the case with the modal share of motorbikes.

Figure 3.3 is a grid-like presentation of trip length distribution (including no-travel) with embedded modal shares and separated for both genders for each of the 20 district categories defined according to size and density (refer table 2.1). Thus it is a compact presentation of a multitude of information which includes size, density, trip length, modal share, gender, and no-travel. It gives us the effect of size and density on trip length and modal shares differentiated by gender. It helps us to consider the extent of gender difference in trip length and modal shares in different urban contexts in India.

3.1 Population Size, Density, and Gendered Work Travel

To say that ‘size matters’ would be a truism. Nonetheless, it is important to understand its extent. One relationship is pretty clear. Generally speaking, as the size of the district reduces, the share of very small trips increases tremendously and the share of medium-long (5-20 km) trips reduces*. There is a clear inversion of trip length distribution as we move from more populated to less populated districts, and this is true for all density categories. Among the districts belonging to very high-density category, this inversion is most obvious. The share of very small trips is more than 40% in the smallest districts and unsurprisingly most (> 90%) of these trips are by walking. In the mid-density districts, reduction in medium-long trips is distributed between very short and short trip lengths. So there is a decent share of 2-5 km long trips in other district groups.

*Note: This is certainly not true for the mega districts (figure 3.1) where the trip patterns do not seem to differ much.


24

This trend is, however, less pronounced among the female workers. But this is only because the percentage of female workers who travel to work is smaller as compared to that for men. Interestingly, significantly more female workers are traveling to work in low-density districts while the same is not true of male workers.

No travelVery small (0-1 km)Small (2-5 km)Medium-long (6-20 km)Very long (> 21 km)

Proportion of Female workers than proportion of male workers (% points)

Figure 3.2: Trip length distribution for women

Figure 3.1: Gender-wise work trip pattern in Indian ‘mega-districts’

F - FemaleM- Male

Den

sity


25

Figure 3.3: Gender-wise work trip pattern in other Indian districts

F - FemaleM- Male

Population

Pop. Density

Very

Hig

hH

igh

Med

ium

Low


26

It cannot be said that women have consistently shorter work trips as compared to men irrespective of population size and population density. Their trip length distribution is similar to men’s in the districts with very high density. But the districts in high and intermediate density categories, where the modal share of two-wheelers is higher, do not show similarity in work trip patterns of males & females. Either these districts have a different employment structure, or there is a lack of accessibility to well-paid employment or both. Figure 3.2 shows the distribution of gender difference for each trip length category. Gender difference is particularly marked out in small trips (2-5 km) and ‘no travel’. Data on the spatial pattern of employment at the district level is not available in India. Similarly, all districts must go through an accessibility evaluation. However, this dimension of lack of accessibility has been given more thought in the following passage while acknowledging that complementary data necessary to make a confident remark is absent.

Prop

ortio

n of

fem

ale

wor

kers

than

pr

opor

tion

of m

ale

wor

kers

(% p

oint

s)

Modal shareϱ = Spearman’s correlation rho***p-value < 0.01

WalkBicycleMTVBus

WalkBicycleMTVBus

ϱ = -0.84***

ϱ = -0.29***

ϱ = 0.34***

ϱ = -0.95***

ϱ = -0.75***

ϱ = -0.24***

ϱ = 0.59***

Small Trips (2-5 km) Medium long trips (6-20km)

Figure 3.4: Gender difference with respect to modal shares of different modes of travel

Across the categories of districts, wherever there is a greater share of the two-wheeler modes (MTW and bicycle), the share of women who travel to work is much lower than that for men (refer to figure 3.4 which shows that there is significant negative correlation between modal share of two-wheeler modes and the proportion of women to men; also refer to figure 3.5 to visualize the all-India gender disparity in the usage of a bicycle for work travel). The trend of lower bicycle ridership among women as compared to men is in harmony with the limited knowledge available on gender-wise modal distribution in Indian cities (for example). This indicates that accessibility to workplaces in such urban contexts is dependent much on two-wheelers which are not accessible to women (either because they don’t know how to ride, or they are not allowed/supposed to ride because of cultural reasons, or they simply do not have the right to utilize them (Astrop, Palmner, Maunder, & Babu, 1996)) and thus hindering the job opportunities which are, at the same time, available to men. This observation is admittedly fragile and a mere conjecture. This also draws attention to the absolute absence of studies probing the causes of gender differences in bicycle ridership in India. Further analysis backed with necessary data can help understand the situation conclusively.

MTW MTW


27

Use of public transport is strikingly more in low-density districts and districts of more than 2 million. Public transport seems to create an impact on women’s decision to get employment at farther locations. There is no data on public transport availability for intra-district (intra-city) travel in Indian districts (or cities).

The higher rate of travel to work among women is clearly attributable to public transport (refer again to figure 3.4) which acts against the relative inaccessibility of two-wheelers to women. This is consistent with findings reported in the literature on gender disparities in work travel. Ironically, the car is also a gender equal mode, and it is equal to the extent that in certain district categories, the proportion of women who drive themselves in the car to work is greater than men. This must not dilute the fact that percentage modal share of cars is hardly a double-digit number in any category (there is a need for further data analysis into the economic background of the car owner and drivers).It must be underlined that walking is the great equalizer, i.e the tendency to walk for women is higher than men. Women, wherever they can and do take to walking, seem to have similar trip length distribution as those for men. Moreover, the percentage of female workers who walk to work is higher as compared to the percentage of pedestrians among the male workers in all of the 24 categories of districts and irrespective of trip length. As discussed in the previous passage, public transport complements walking by providing a greater range of opportunities. Walking among the females is higher than the males for short as well as very short trips. Apart from the hill states, walking to work is more common in parts of central India (see figure 3.2).

3.2 Gender and the ‘No-travel’

As discussed in the first chapter, obscurity around the data on self-employment and location of the workplace has made it difficult to understand the dynamics of not traveling to work. Therefore, the limited gender insight available to us becomes precious. Figure 3.5 presents the box plot of percentage of ‘no-travel’ across all the 640 districts. The median percentage of no-travel for women is nearly 15% age points higher than that for men. If outliers are ignored, no-travel among female shows greater variation across districts. In the context of what has been discussed in the previous section, this is an indication for a greater role of urban environment in shaping women’s travel as compared to that of men.

Male Female

Figure 3.5: Comparing the ‘no travel’ among male and female workers

What are these outliers? All the outlier districts for male belong to hill areas of J&K, Arunachal Pradesh, Assam, Manipur, Nagaland, Meghalaya. All, except one, outlier districts for the female belong to Uttar Pradesh. There are 26 out of 640 districts where the percentage of workers not traveling among men is at least 5% higher than that among women. Only one of these districts (Bijapur, Maharashtra) is located in a non-hilly terrain. Districts in Uttar Pradesh, Bihar, and Rajasthan among the states have highest gender difference in the rate of no-travel. District-level gender differentials among other states (excluding hill states) is similar. District-level difference in no-travel between female and male workers (percentage points) is shown in figure 3.7. Again, there is remarkable variation in gender difference in commute to work across the different parts of the country. Traveling to work is less among the women


28

Figure 3.6 (a): Distance-wise percentages of the male and female workers who walk to work

Figure 3.6 (b): Percentage of male or female workers who ride a bicycle to workplace

Male Female

Wom

enM

ale


29

District in hill state

Districts in Rajasthan, UP, Bihar (left to right)

Districts in other states

(% n

o-tr

avel

in w

omen

) – (%

no-

trav

el in

men

)

State codeFigure 3.7: Percentage of no-travel workers among the women

minus that among the men

The disparity index proposed by David Sopher,which formulates disparity as the log of odds ratio (Sopher, 1974), can be used to create a gender disparity index for no-travel. If the percentage of women workers not traveling to work is p and the percentage of male workers not traveling to work is q then odds of women workers not traveling would be p/(100-p) and the odds of male workers not traveling would be q/(100-q). In this case, Sopher’s disparity index (DIS) for no-travel will become-

DIS = log (p/q) + log [(100-q)/(100-p)]

Figure 3.9 shows the spatial variation of DIS of ‘no-travel’ across the all the districts. DIS equal to 0 means perfect parity in the tendency of not traveling. A higher positive value of DIS would mean the higher tendency of no-travel among the women whereas the male workers would have a greater tendency of not traveling to work in a region with negative DIS. Positive disparity (i.e. higher tendency of no-travel among women than men) is significantly higher in UP, Bihar and Rajasthan and Gujarat, but not so much in the central peninsular region. But in most parts of the country, except in the hill states and a part of Chhattisgarh, the disparity is on the positive side. Pearson correlation coefficient for no-travel in male and female workers is significant at p < 0.001 and has a value of 0.63. Locations with high proportion of no travel to work amongst male also have high proportions no-travel to work for female. But having said that, the female workers have generally a higher percentage of no-travel than is the case with male workers. However, the relationship varies regionally. Another important highlight is that no-travel is not affected as much by population size as by population density. Also, while

No travel (Male)

No Travel(Female)

Population Size

Population Density

Bicycle Ownership

MTW Ownership

Car Ownership

No Travel (Male)

1.00 0.63* -0.26* 0.06 -0.30* -0.45* 0.02

No Travel (Female) 1.00 -0.12** 0.17* 0.17* -0.26* -0.28*

Population Size 1.00 0.18* 0.012 0.24* 0.16*

Population Density 1.00 0.01 0.10 0.18*

Bicycle Ownership 1.00 0.41* -0.30*

MTW Ownership 1.00 0.33

Car Ownership 1.00

Table 3.1 Correlation analysis of no-travel and other variables at the district level

workers of UP, Bihar, Rajasthan and the central peninsular region which includes parts of Maharashtra, Karnataka, Telangana and Andhra Pradesh.

Note: *p < 0.001; **p < 0.01


30

Figure 3.8: Spatial distribution of no-travel workers in urban India

Figure 3.9: Spatial variation in Sopher’s Gender Disparity Index (DIS) of ‘no-travel’

density is positively correlated (r= 0.17, p < 0.001) with no-travel among female, size has a weak negative correlation (r= -0.12, p < 0.005) with the same. Table 3.1 presents the correlation analysis of district-level characteristics for which the data is available. Before closing the discussion on no-travel, it is important to make it clear that no-travel does not necessarily mean immobility. If the work available at home or in the neighborhood is meaningful and pays well, no-travel is indeed a positive scenario. But, if the economic circumstances force the women to sell their labor but access to a good job remains restricted, lack of mobility caused due to unchanging household roles and employment disparities (Rosenbloom, 2004) cannot be overlooked.


31

SUGGESTIONS

Source: Unsplash


32

When it comes to transportation planning, the importance of quantitative data on travel choices of the people can hardly be underestimated. Such data is essentially the backbone of transportation demand modeling irrespective of the analytical approach adopted. In this respect, the initiative to collect census data on work trips is definitely a welcome step because it provides a comprehensive insight, even if only at an aggregated level. As has been pointed out in the beginning of this report the data collected in the Census 2011 is limited. It provokes more questions than it gives answers. First, it concentrates solely on the travel to work, and that too by the workers falling under the category of ‘other workers’. Second, it is not clear how the information on distance and mode of travel were collected in case of the vendors, delivery persons and other workers with an irregular place of work. Third, even for the reported work trips, the information is limited to the mode and (approximate) travel distance. Reliability of calculation, or rather estimation, of travel distance, is another limitation in itself. Therefore, this report ends with few comments on how the census dataset on travel can be improved so that it helps the urban transport planners in India. The suggestions can be grouped into two categories: in terms of the scope of the survey, and in terms of quality of the data collected.

Scope of the survey

• The first set of suggestions is related to extending the scope of the census questionnaire to collect data on travel pattern of each household. Transportation demand models require the household-level information as the household is the place of origin of all the trips made by the members of a household. Depending on the household demographics, information about the regular trips undertaken by different members for various purposes should also be included.

• A question on available modes of transport should necessarily be added to the questionnaire. This would not only enable better appreciation of the behavioral aspect of regular travel in India but it will also help align transportation planning more closely with economic policymaking, especially as it relates to the creation and provision of employment (Green, 1995). Data on availability of modes at a disaggregated level is also essential to frame the existing paradigm of mobility and to better plan the transition to a more sustainable paradigm.

• Census has been recording the availability of vehicles for personal use for each household. Vehicle ownership data should also include the number of each type of vehicles available for the member(s) of the household.

• Good data on work travel must also make it possible to understand the inaccessibility to job-locations faced by different population sub-groups. The census should ask the workers, including the marginal workers and unemployed workforce, about the places where the respondents are seeking jobs and the places which they are avoiding. A simplistic indicator for this kind of information can be the maximum time and a distance respondent would be prepared to travel if a more satisfying job is made available.

Data quality and reliability

The Census data on work travel has been published at the district level. However, the data must be prepared and published at the sub-district/ward level or the household level for a more meaningful and direct use in transport planning. As has been shown in this report, the census data reveals that residence-based employment has a significant proportion of all


33

employees, and the workers are largely traveling smaller distances. Under such circumstances, publishing the data only at ward level will make it possible to make complete sense of linkages between travel decisions and local labor markets and economy. Availability of data at the smallest possible scale is necessary also because it reveals the heterogeneity of travel behavior within the districts and the sub-districts. Census data on travel-to-work, if made available at the ward-level, can also be used to create pollution emission profiles thus empowering the air quality management interventions.

A big question mark looms over the reliability of distance (actually estimated distance) recorded in the Census 2011. First, it is not clear if the travels of workers with ‘no fixed place of work’ such as vendors and casual laborers have been recorded as ‘no travel’. Second, there are some districts which are very small in area but with quite a significant share of trips involving very long distances. Third, accurate estimation of distance to work place is not a simple task on the part of the respondents and it seems intuitive that respondents might have guessed the distances. Therefore the question on distance must be complimented with a question on travel time. Another option is to record the place of work or the ward details of the workplace location.

It is not unusual of work trips and other kinds of trips to involve multiple modes of transport. Census 2011 instrument allowed only a single mode to be recorded per individual. The multi-modality of trips, if it exists in case of an individual’s regular travel, must be accommodated and should be duly recorded in the survey. The modal choice set used by the Census does not make it clear which mode category the cycle-rickshaw is put in. It is certainly problematic to group them with the para-transit modes such as auto-rickshaw.

Currently there is no organization in India which regularly collects data on household travel. In the absence of such an agency, it is only the census data that can responsibly provide a comprehensive idea of travel patterns in India. To summarize the suggestions, here is the list of ten recommendations to the Census of India by the authors of this report:

1. To facilitate the applicability in transportation planning, the scope of the survey can be extended to record the most regular trips of each member of a household.

2. A question on available modes should be added in the survey instrument to better understand the mode choice decisions and the interventions needed to make transition to more sustainable transport system.

3. Vehicle availability should also include the number of vehicles of each kind available for use.

4. To understand the status of job accessibility, a question can added to record the maximum time and distance a respondent would be prepared to travel if a more satisfying job is made available.

5. Data must be published at household level, or at least at the ward-level for a more meaningful use in transportation planning.

6. Reliability of estimations of travel distances should be improved. An accurate estimation of distance to work place is not a simple task on part of the respondents. Therefore the question on distance must be complimented with a question on travel time and the mention of the actual work place location. Both travel time and distance have been shown to have a lack of reliability; in certain cases, we should also try to validate the data for a smaller subsample.


34

7. Place of work or the ward details of the workplace location can also be recorded. 8. It should be made clear if the travels of workers with ‘no fixed place of work’ such as

vendors and casual laborers have been recorded as ‘no travel’.9. Multi-modality of trips, if it exists in case of an individual’s regular travel, must be

accommodated and should be duly recorded in the survey.10. The modal choice set used by Census does not make it clear which mode category the

cycle-rickshaw is put in. Mode categories must be rationalized.


35

REFERENCES / APPENDIX

Source: Unsplash


36

REFERENCES:Astrop, A., Palmner, C., Maunder, D., & Babu, D. (1996). The urban travel behaviour and constraints of low income households and females in Pune , India. In National Conference on Women’s Travel Issues (pp. 212–246). Baltimore, Maryland. Retrieved from https://www.fhwa.dot.gov/ohim/womens/chap12.pdfGreen, A. E. (1995). Using census and survey commuting statistics in local labour market analysis. Local Economy, 10(3), 259–273. https://doi.org/10.1080/02690949508726287Rosenbloom, S. (2004). Understanding Women’s and Men’s Travel Patterns. In Research on Women’s Issues in Transportation (pp. 12–13). Chicago, Illinois: Transportation Research Board.Sopher, D. E. (1974). A Measure of Disparity. The Professional Geographer, 26(4), 389–392.

R packages usedAdrian A. Dragulescu (2014). xlsx: Read, write, format Excel 2007 and Excel 97/2000/XP/2003 files. R package version 0.5.7. https://CRAN.R-project.org/package=xlsx

Baptiste Auguie (2016). gridExtra: Miscellaneous Functions for "Grid" Graphics. R package version 2.2.1. https://CRAN.R-project.org/package=gridExtra

H. Wickham. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2009.

Hadley Wickham and Romain Francois (2016). dplyr: A Grammar of Data Manipulation. R package version 0.5.0. https://CRAN.R-project.org/package=dplyr

Jakson Aquino. Includes R source code and/or documentation written by Dirk Enzmann, Marc Schwartz, Nitin Jain and Stefan Kraft (2016). descr: Descriptive Statistics. R package version 1.1.3. https://CRAN.R-project.org/package=descr

Original S code by Richard A. Becker, Allan R. Wilks. R version by Ray Brownrigg. Enhancements by Thomas P Minka and Alex Deckmyn. (2017). maps: Draw Geographical Maps. R package version 3.2.0. https://CRAN.R-project.org/package=maps

Pebesma, E.J., R.S. Bivand, 2005. Classes and methods for spatial data in R. R News 5 (2). https://cran.r-project.org/doc/Rnews/.

R Core Team (2016). foreign: Read Data Stored by Minitab, S, SAS, SPSS, Stat, Systat, Weka, dBase, .... R package version 0.8-67. https://CRAN.R-project.org/package=foreign

Robert J. Hijmans (2016). raster: Geographic Data Analysis and Modeling. R package version 2.5-8. https://CRAN.R-project.org/package=raster

Roger Bivand and Nicholas Lewin-Koh (2017). maptools: Tools for Reading and Handling Spatial Objects. R package version 0.9-2. https://CRAN.R-project.org/package=maptools

Roger Bivand, Tim Keitt and Barry Rowlingson (2017). rgdal: Bindings for the Geospatial Data Abstraction Library. R package version 1.2-8. https://CRAN.R-project.org/package=rgdal

Roger S. Bivand, Edzer Pebesma, Virgilio Gomez-Rubio, 2013. Applied spatial data analysis with R, Second edition. Springer, NY. http://www.asdar-book.org/ Ryan M. Hope (2013). Rmisc: Rmisc: Ryan Miscellaneous. R package version 1.5.https://CRAN.R-project.org/package=Rmisc


37

Appendix I: Definitions

• Urban areaUrban areas are comprised of two types of administrative units – Statutory

• Towns and Census Towns.1. Statutory Towns: All administrative units that have been defined by statute as urban like

municipal Corporation, Municipality, Cantonment Board, Notified Town Area Committee, Town Panchayat, Nagar Palika etc., are known as Statutory Towns.

2. Census Towns: Administrative units satisfying the following three criteria simultaneously are treated as Census Towns:

i) A minimum population of 5,000 persons *; ii) 75% and above of the male main working population being engaged in non–agricultural pursuits; and iii) A density of population of at least 400 persons per sq. km. (1,000 per sq.mile). *For the purpose of identification of places that qualify to be classified as3. ‘Census Towns’, all villages with a population of 4000 and above as per the Census 2001, a

population density of 400 persons per sq. km. and having at least 75 per cent of male main working population engaged in non– agricultural activity were considered.

• CityTowns with population of 1,00,000 and above are categorized as cities.

• Out Growth and Urban Agglomeration An Out Growth (OG) is a viable unit such as a village or a hamlet or an enumeration block made up of such village or hamlet and clearly identifiable in terms of its boundaries and location. Some of the examples are railway colony, university campus, port area, military camp, etc., which have come up near a statutory town outside its statutory limits but within the revenue limits of a village or villages contiguous to the town. While determining the outgrowth of a town, it has been ensured that it possesses the urban features in terms of infrastructure and amenities such as pucca roads, electricity, taps, drainage system for disposal of waste water etc., educational institutions, post offices, medical facilities, banks etc. and physically contiguous with the core town of the UA. Each such town together with its outgrowth(s) is treated as an integrated urban area and is designated as an ‘urban agglomeration’. An urban agglomeration is a continuous urban spread constitutinga town and its adjoining outgrowths (OGs), or two or more physically contiguous towns together with or without outgrowths of such towns. An Urban Agglomeration must consist of at least a statutory town and its total population (i.e. all the constituents put together) should not be less than 20,000 as per the 2001 Census. In varying local conditions, there were similar other combinations which have been treated as urban agglomerations satisfying the basic condition of contiguity. Examples: Greater Mumbai UA, Delhi UA, etc.In the 2011 Census, 475 places with 981 OGs have been identified as Urban Agglomerations.

• WorkWork is defined as participation in any economically productive activity with or without compensation, wages or profit. Such participation may be physical and/or mental in nature. Work involves not only actual work but also includes effective supervision and direction of work. It even includes part time help or unpaid work on farm, family enterprise or in any other economic activity. All persons (irrespective of age and sex) who participated


38

in any economically productive activity for any length of time during the reference period are defined as workers. Normally, production for self-consumption is not treated as economic activity. However, for the purpose of census an exception has been made in the case of persons who are engaged in growing of crops (except plantation crops), rearing of animals and milk production for self-consumption. These categories have been treated as 16 economic activity. Reference period for determining a person as worker and non-worker is one year preceding the date of enumeration.

1. Main Workers Workers who worked for more than 6 months (180 days) in the reference period are termed as Main Workers.2. Marginal Workers Workers who worked for less than six months (180 days) in the reference period are termed as Marginal Workers. Marginal workers are further bifurcated into two categories i.e. those who worked for 3 months or more but less than 6 months and those who worked for less than 3 months.3. Cultivators

For purpose of the Census, a person is classified as cultivator if he or she is engaged in cultivation of land owned or held from Government or held from private persons or institutions for payment in money, kind or share. Cultivation includes effective supervision or direction in cultivation. A person who has given out her/his land to another person or persons or institution(s) for cultivation for money, kind or share of crop and who does not even supervise or direct cultivate on land, is not treated as cultivator.

4. Agricultural LabourersA person who works on another person’s land for wages in money or kind or share is regarded as an agricultural labourer. She or he has no risk in the cultivation, but merely works on another person’s land for wages. An agricultural labourer has no right of lease or contract on land on which she/he works.

5. Household Industry WorkersHousehold Industry is defined as an industry conducted by one or more members of the household at home or within the village in rural areas and only within the precincts of the house where the household lives in urban areas. The larger proportion of workers in the household industry consists of members of the household. The industry is not run on the scale of a registered factory where more than 10 persons with power or 20 persons without power is in use as it would qualify or has to be registered under the Indian Factories Act.The main criterion of a Household industry even in urban areas is the participation of one or more members of a household. Even if the industry is not actually located at home in rural areas there is a greater possibility of the members of the household participating even if it is located anywhere within the village limits. In the urban areas, where organized 17 industry takes greater prominence, the Household Industry should be confined to the precincts of the house where the participants live.

6. Other WorkersWorkers other than cultivators, agricultural labourers or workers in Household Industry, as defined above are termed as ‘Other Workers’ (OW). Examples of such type of workers are government servants, municipal employees, teachers, factory workers, plantation workers, those engaged in trade, commerce, business, transport, banking, mining, construction, political or social work, priests, entertainment artists, etc.


39

Appendix II Classification of UA/Towns

Class I: The UAs/Towns which have at least 100000 persons as population are categorized as Class I UA/Town. At the Census 2011, there are 468 such UAs/Towns. The corresponding number in Census 2001 was 394.Million-plus UAs/Towns: Out of 468 UAs/Towns belonging to Class I category, 53UAs/Towns each has a population of one million or above each. Known as MillionPlus UAs/Cities, these are the major urban centres in the country. 160.7 millionpersons (or 42.6% of the urban population) live in these Million-plus UAs/Cities.Mega Cities: Among the Million-plus UAs/Cities, there are three very large UAs withmore than 10 million persons in the country, known as Mega Cities. These areGreater Mumbai UA (18.4 million), Delhi UA (16.3 million) and Kolkata UA (14.1million).****No classification brackets for districts in census**** This report therefore uses the same bracket criteria for districts as used to classify cities by Census of India.

Table A-1: Number of districts belonging to each population category

Population Category Number of districts

1 to 5 Lakh 285

5 to 10 Lakh 33

10 to 20 Lakh 59

20 to 40 Lakh 32

40 to 80 Lakh 06

More than 80 Lakh 03

Total number of 1 Lakh+ districts = 418Total number of million+ districts = 100

Appendix III Population density distribution

These charts are shown to justify the intervals selected for classifying districts in terms of population density. There are very few but a significant number of districts with a density of more than 10000 people per square kilometer. Similarly, another major break is at 5000 ppsk

Figure A-1: Bivariate distribution of population and population density (aggregated at district level)


40

(Figure A-2). In the left of the two charts in figure A-2, 2000 and 5000 ppsk can be viewed as the two breaking points. Therefore the classification scheme selected in the study is- low density (0-2000 persons per sq. km), moderate density (2-5000 persons per sq. km), high density (5-10000 persons per sq. km), and very high density (above 10000 persons per sq. km).

Figure A-2: Histogram of population density in India (densities aggregated at district level)

Appendix IV Regions and Regional Variation

Table A-2: List of regions- grouping the neighboring states

S. No. Region States

1 North India Jammu & Kashmir, Himachal Pradesh, Punjab, Chandigarh, Uttarakhand, Haryana, NCT of Delhi, Rajasthan, Uttar Pradesh

2 North Eastern States Sikkim, Arunachal Pradesh, Nagaland, Manipur, Mizoram, Tripura, Meghalaya, Assam

3 East India Bihar, Jharkhand, West Bengal, Odisha

4 Central India Chhattisgarh, Madhya Pradesh

5 West India Gujarat, Daman & Diu, Dadra & Nagar Haveli, Maharashtra

6 South India Andhra Pradesh, Karnataka, Goa, Kerala, Tamil Nadu, Puducherry

7 Islands of India Lakshadweep, Andaman & Nicobar Islands

Table A-3: Trip length distribution of work trips in different regions of India

S. No

Name No travel

Up to 1 km

2-5 km 6-10 km

11-20 km

21-30 km

31-50 km

> 50 km

1 North India Region 35.8 15.0 21.1 13.1 6.5 3.7 2.1 2.6

2 North East India 33.1 24.8 23.3 9.7 3.5 2.1 1.4 2.2

3 East India 31.6 17.9 22.6 13.1 5.5 3.9 2.3 3.2

4 Central India Region 29.1 19.2 27.4 12.8 5.2 2.8 1.7 1.7

5 West India Region 25.4 16.6 25.6 14.7 7.7 4.3 3.3 2.3

6 South India Region 28.0 15.7 23.5 15.6 8.1 4.0 2.8 2.9

7 Islands of India 20.5 26.8 29.8 14.6 5.2 1.6 0.7 0.6


41

Table A-4: Mode share of work trips in different regions of India

NMT PERSONAL MV PARA-TRANSIT

PUBLIC TRANSPORT

S. No. Name Walk Bicycle MTW Car IPT Bus Train Others

1 North India Region 32.5 22.1 18.4 4.9 3.9 13.5 3.5 1.3

2 North East India 47.0 22.0 8.6 4.4 5.0 10.3 1.2 1.4

3 East India 35.0 30.1 10.4 2.2 2.7 9.6 8.5 1.4

4 Central India 38.2 24.4 22.4 2.2 2.4 7.5 2.1 0.7

5 West India Region 29.6 12.6 23.8 4.0 7.1 11.8 9.9 1.2

6 South India Region 29.5 11.9 19.2 4.0 4.4 27.5 2.2 1.3

7 Islands of India 34.8 11.7 18.3 4.8 2.5 23.7 0.1 4.0

Appendix V Datasheets on states

Figure A-3: Poverty levels ( percentage of population below poverty line) vs. area of state


42

Table A-5: States and UTs in descending order of urbanization

S. No. State/UT Share of urban population (%) Rate of urbanization (%)*

1 Delhi 97.5 126.6

2 Chandigarh 97.2 126.9

3 Lakshadweep 78.1 186.6

4 Daman & Diu 75.2 318.4

5 Puducherry 68.3 131.1

6 Goa 62.2 135.2

7 Mizoram 51.5 127.4

8 Tamil Nadu 48.4 127.2

9 Kerala 47.7 192.7

10 Dadra & Nagar Haveli 46.6 316.7

11 Maharashtra 45.2 123.7

12 Gujarat 42.6 135.8

13 Karnataka 38.6 131.3

14 Punjab 37.5 125.7

15 Andaman & Nicobar Islands

35.7 116.6

16 Haryana 34.8 144.3

17 Andhra Pradesh 33.5 136.3

18 West Bengal 31.9 129.9

19 Uttarakhand 30.6 141.9

20 Manipur 30.2 142.7

21 Nagaland 29.0 167.4

22 Madhya Pradesh 27.6 125.6

23 Jammu & Kashmir 27.2 135.7

24 Tripura 26.2 176.1

25 Sikkim 25.0 253.4

26 Rajasthan 24.9 129.3

27 Jharkhand 24.1 132.3

28 Chhattisgarh 23.2 141.8

29 Arunachal Pradesh 22.7 137.5

30 Uttar Pradesh 22.3 128.8

31 Meghalaya 20.1 131.0

32 Odisha 16.7 126.8

33 Assam 14.1 127.6

34 Bihar 11.3 135.1

35 Himachal Pradesh 10.0 115.6


43

Table A-6: Highway road density in states (descending order)

State Highway density (per 100 sq. km)

Uttarakhand 29.3

Kerala 14.9

Goa 14.8

Karnataka 13.1

Tamil Nadu 12.1

Maharashtra 12.0

Gujarat 11.5

Manipur 11.0

Tripura 10.4

Bihar 9.5

Haryana 9.4

Meghalaya 9.0

Nagaland 9.0

Mizoram 8.2

West Bengal 8.1

Puducherry 7.9

Assam 7.7

Punjab 6.0

Himachal Pradesh 5.6

Chhattisgarh 5.6

Andhra Pradesh 5.5

Rajasthan 5.1

Jharkhand 5.1

Madhya Pradesh 5.0

Odisha 4.7

Sikkim 4.6

Andaman & Nicobar Islands

3.0

Arunachal Pradesh 2.4

Uttar Pradesh 2.4

Jammu & Kashmir 0.6

Chandigarh -

D & N Haveli -

Daman & Diu -

Delhi -

Lakshadweep -

Table A-7: Urban poverty in states (ascending order)

States Urban Poor Population (2011-12)

(%)

Andaman & Nicobar Islands 0.00

Lakshadweep 3.44

Sikkim 3.66

Goa 4.09

Himachal Pradesh 4.33

Kerala 4.97

Andhra Pradesh 5.81

Puducherry 6.30

Mizoram 6.36

Tamil Nadu 6.54

Jammu & Kashmir 7.20

Tripura 7.42

Maharashtra 9.12

Punjab 9.24

Meghalaya 9.26

Delhi 9.84

Gujarat 10.14

Haryana 10.28

Uttarakhand 10.48

Rajasthan 10.69

Daman & Diu 12.62

West Bengal 14.66

Karnataka 15.25

Dadra & Nagar Haveli 15.38

Nagaland 16.48

Orissa 17.29

Arunachal Pradesh 20.33

Assam 20.49

Madhya Pradesh 21.00

Chandigarh 22.31

Chhattisgarh 24.75

Jharkhand 24.83

Uttar Pradesh 26.06

Bihar 31.23

Manipur 32.59

All India 13.70


44

Table A-8: Per capita GSDP of Indian States/UTs (at constant prices)*

Sl. State\UT GSDP (2013-14) (cr. INR) Per Capita GSDP (INR)

1 Goa 30345 208167.1

2 Chandigarh 15688 148745.7

3 Delhi 219991 131312.5

4 Puducherry 14077 113117.0

5 Andaman & Nicobar Islands

4220 111069.0

6 Sikkim 6152 101236.2

7 Maharashtra 896767 79802.7

8 Haryana 199657 78750.6

9 Gujarat 452625 74958.2

10 Uttarakhand 70926 70107.5

11 Himachal Pradesh 47255 68919.9

12 Kerala 226208 67751.9

13 Tamil Nadu 480618 66623.9

14 Punjab 174038 62820.0

15 Nagaland 11367 57391.6

16 Andhra Pradesh 453151 53522.5

17 Karnataka 321455 52584.9

18 Mizoram 5608 51401.7

19 Tripura 18732 51026.5

20 Meghalaya 13347 45030.2

21 Arunachal Pradesh 5905 42709.0

22 West Bengal 371795 40701.0

23 Rajasthan 257432 37515.0

24 Chhattisgarh 95262 37298.8

25 Jammu & Kashmir 45847 36534.6

26 Jharkhand 109408 33187.9

27 Odisha 137468 32771.5

28 Madhya Pradesh 230095 31694.6

29 Manipur 8330 30605.2

30 Assam 86862 27867.8

31 Uttar Pradesh 464510 23274.2

32 Bihar 173409 16705.3*Base year 2004-05


45

Table A-9: State-wise percentage of workers in different trip length categories

S. No. State No travel Up to 1 km

2-5 km 6-10 km

11-20 km

21-30 km

31-50 km

> 50 km

1 JAMMU & KASHMIR 46.1 11.1 15.2 13.3 5.7 2.4 2.3 3.8

2 HIMACHAL PRADESH

29.3 19.2 22.8 14.3 6.4 2.8 2.4 2.9

3 PUNJAB 31.9 15.5 24.2 15.0 6.2 3.6 1.6 1.9

4 CHANDIGARH 15.5 12.3 33.5 26.6 7.9 1.7 1.2 1.2

5 UTTARAKHAND 31.0 19.2 24.5 13.1 5.4 2.7 1.7 2.2

6 HARYANA 31.1 15.6 21.8 13.1 6.6 3.9 3.5 4.4

7 NCT OF DELHI 15.9 16.1 24.5 18.4 15.0 6.3 3.0 0.8

8 RAJASTHAN 32.0 16.4 24.2 12.4 6.1 2.9 2.5 3.5

9 UTTAR PRADESH 44.2 13.7 18.0 11.5 4.9 3.8 1.6 2.4

10 BIHAR 44.1 15.2 19.2 11.4 3.5 3.1 1.2 2.2

11 SIKKIM 39.3 24.9 19.6 9.3 3.1 1.5 1.2 1.2

12 ARUNACHAL PRADESH

37.7 30.3 16.8 9.7 2.5 1.1 0.7 1.1

13 NAGALAND 43.6 27.8 17.3 6.9 1.9 1.0 0.6 1.0

14 MANIPUR 46.4 15.2 17.2 10.3 4.4 3.1 1.9 1.5

15 MIZORAM 33.2 37.0 20.2 6.0 1.3 0.4 0.5 1.3

16 TRIPURA 25.9 23.6 29.5 11.8 4.2 2.1 1.3 1.6

17 MEGHALAYA 36.5 23.4 21.1 9.7 2.9 1.5 1.8 3.1

18 ASSAM 31.5 25.2 24.1 9.6 3.5 2.2 1.4 2.4

19 WEST BENGAL 29.2 18.8 21.5 12.9 6.1 4.3 3.0 4.1

20 JHARKHAND 25.0 18.1 28.9 15.3 5.7 3.4 1.5 2.1

21 ODISHA 26.8 18.3 25.8 13.9 6.2 4.4 2.0 2.6

22 CHHATTISGARH 22.6 21.1 30.6 13.1 6.2 3.0 1.8 1.6

23 MADHYA PRADESH 31.3 18.5 26.4 12.7 4.9 2.7 1.7 1.8

24 GUJARAT 27.4 16.5 29.0 14.3 6.1 3.2 1.8 1.6

25 DAMAN & DIU 7.9 45.8 31.6 9.4 1.6 0.5 0.2 3.0

26 DADRA & NAGAR HAVELI

15.9 34.8 26.2 14.6 5.2 1.8 1.0 0.6

27 MAHARASHTRA 24.5 16.4 23.7 15.0 8.6 4.9 4.2 2.6

28 ANDHRA PRADESH 35.9 14.4 21.3 13.1 6.6 3.8 2.3 2.5

29 KARNATAKA 29.5 15.9 21.7 15.5 8.3 3.8 2.7 2.6

30 GOA 23.6 13.3 24.4 16.1 10.9 5.5 4.8 1.6

31 LAKSHADWEEP 20.1 50.5 26.9 1.6 0.0 0.6 0.0 0.2

32 KERALA 24.2 17.1 25.8 17.1 7.8 3.2 2.2 2.6

33 TAMIL NADU 22.4 15.6 24.7 16.4 9.1 4.7 3.4 3.7

34 PUDUCHERRY 11.2 15.7 34.1 20.5 9.7 4.5 1.7 2.5

35 ANDAMAN & NICOBAR ISLANDS

20.6 22.9 30.3 16.8 6.1 1.8 0.8 0.7


46

Table A-10: State-wise percentage of work trips in different travel mode categories

NMT Personal Para-transit Public Transport

S. No. Name Walk Bicycle MTW Car IPT Bus Train Others

1 JAMMU & KASHMIR 39.2 3.1 6.6 6.1 2.8 39.1 1.5 1.5

2 HIMACHAL PRADESH 49.2 2.8 7.9 4.0 1.1 34.0 0.5 0.5

3 PUNJAB 26.5 29.0 24.8 4.3 3.1 9.6 1.5 1.3

4 CHANDIGARH 16.3 30.4 25.3 14.6 3.5 7.9 0.4 1.7

5 UTTARAKHAND 41.0 17.0 18.2 5.4 4.6 11.2 1.5 1.2

6 HARYANA 29.8 18.6 19.2 7.5 5.5 13.6 4.7 1.2

7 NCT OF DELHI 26.3 10.7 16.7 12.8 2.7 25.9 3.5 1.4

8 RAJASTHAN 37.2 13.1 22.6 3.6 3.9 16.2 2.1 1.2

9 UTTAR PRADESH 31.9 31.1 16.7 3.0 4.3 6.8 4.9 1.3

10 BIHAR 40.7 28.9 11.3 1.8 3.6 5.4 7.2 1.1

11 SIKKIM 62.4 0.5 1.5 21.1 9.5 3.5 0.4 1.1

12 ARUNACHAL PRADESH 65.2 5.5 9.0 10.0 2.2 4.4 0.5 3.1

13 NAGALAND 59.9 3.4 5.4 11.4 10.0 8.1 0.4 1.5

14 MANIPUR 32.6 14.9 17.1 5.6 11.4 14.0 1.1 3.4

15 MIZORAM 58.1 1.9 11.0 6.8 5.9 14.9 0.3 1.2

16 TRIPURA 46.1 25.0 7.9 4.4 7.2 7.8 0.8 0.9

17 MEGHALAYA 54.7 2.2 3.9 13.0 13.6 10.5 0.3 1.8

18 ASSAM 45.4 27.2 8.6 2.4 3.0 10.7 1.5 1.2

19 WEST BENGAL 32.0 30.2 5.7 2.6 1.8 13.4 12.6 1.7

20 JHARKHAND 38.0 28.6 16.7 2.2 5.3 5.1 3.4 0.7

21 ODISHA 33.9 32.8 18.1 1.7 1.9 7.6 2.4 1.5

22 CHHATTISGARH 31.6 32.6 24.6 2.1 1.7 4.4 2.2 0.7

23 MADHYA PRADESH 40.4 21.6 21.7 2.3 2.7 8.5 2.1 0.7

24 GUJARAT 28.1 16.1 30.9 4.0 10.1 7.9 1.7 1.3

25 DAMAN & DIU 62.5 9.9 14.2 1.7 3.5 6.3 0.6 1.2

26 DADRA & NAGAR HAVELI

46.0 11.4 17.8 3.4 9.3 11.4 0.4 0.3

27 MAHARASHTRA 30.2 10.7 20.0 4.0 5.5 14.0 14.5 1.2

28 ANDHRA PRADESH 30.6 16.3 19.8 2.8 8.4 18.2 2.1 1.9

29 KARNATAKA 33.6 8.0 18.2 5.7 4.7 27.2 1.5 1.0

30 GOA 19.8 3.1 31.2 7.3 2.1 34.6 0.6 1.3

31 LAKSHADWEEP 36.6 41.0 17.1 0.4 0.4 0.3 0.2 4.0

32 KERALA 33.0 5.1 13.8 3.9 3.1 37.4 2.6 1.2

33 TAMIL NADU 24.1 15.0 21.7 3.6 1.7 29.9 2.8 1.2

34 PUDUCHERRY 15.6 20.3 37.7 2.7 2.3 18.3 0.5 2.6

35 ANDAMAN & NICOBAR ISLANDS

34.5 6.8 18.5 5.5 2.9 27.6 0.1 4.0


47

Appendix VI Census survey questions on travel to workplace

Census Metadata provides the question asked to respondents. It mentions the following: This question was asked only for the “Other workers” i.e. other than cultivators, agricultural labourers and household industry workers. In addition, this question was also not applicable for the defence forces and similar paramilitary personnel. If the person was engaged in more than one economic activities during the last year, this question was asked with reference to the main economic activity. This question has two parts:(a) One way distance from residence to place of work in kilometers: In case the journey from the residence to place of work is carried out through any mode of land transport, the road distance was recorded. However, if the journey was performed by any modes relating to water transport, then the aerial distance was recorded.(b) Mode of travel to place of work: The mode of travel i.e. how the distance from the residence was covered by the person to reach her/his place of work was ascertained and appropriate code was recorded. The codes used were: On foot-1, Bicycle-2, Moped/Scooter/Motor cycle-3, Car/Jeep/Van-4, Tempo/Auto rickshaw/Taxi-5, Bus-6, Train-7, Water transport-8, Any other-9, No travel-0.

Appendix VII Work location in NSS data (68th round)

Categories used by NSS to record the location of workplace were following (separately for rural and urban workplace locations):

Table A-11: Location of workplace categories used in NSS 68th round]

S. No. Workplace located in percentage of workers

(only urban workplaces)

1 Own dwelling unit 3.94

2 Structure attached to own dwelling unit

4.13

3 Open area adjacent to own dwelling unit*

-

4 Detached structure adjacent to own dwelling unit

0.79

5 Own enterprise/unit/office/shop but away from own dwelling

5.12

6 Employer’s dwelling unit 4.13

7 Employer’s enterprise/unit/office/shop but outside employer’s dwelling

69.49

8 Street with fixed location 2.17

9 Construction site 3.35

10 Others 0.98

11 No fixed workplace 5.91*invalid category for urban workplaces


48

Table A-12: Classification of workers in different occupational divisions based on NCO 2005

Division Code Name of occupational division percentage of workers

1 Legislators, senior officials and managers 4.4

2 Professionals 6.4

3 Technicians and associate professionals 6.8

4 Clerks 6.2

5 Service workers and shop & market sales workers

21.7

6 Skilled agricultural and fishery workers 1.4

7 Craft and related trades workers 21.3

8 Plant and machine operators and assemblers

9.6

9 Elementary occupations 17

10 Unclassified 5.2

Appendix VIII State-wise shares of establishment structure (6th Economic Survey)

Table A-13: State-wise distribution of establishment structure*

percentage of all establishments

State Outside household without fixed structure

Inside household percentage of ‘No- travel’ workers

1 Jammu & Kashmir 9.7 11.0 46.1

2 Himachal Pradesh 8.2 23.3 29.3

3 Punjab 14.4 16.7 31.9

4 Chandigarh 58.7 7.6 15.5

5 Uttarakhand 10.7 18.0 31.0

6 Haryana 10.4 17.5 31.1

7 Delhi 21.2 24.9 15.9

8 Rajasthan 9.8 18.0 32.0

9 Uttar Pradesh 14.3 33.3 44.2

10 Bihar 8.1 22.6 44.1

11 Sikkim 11.0 40.8 39.3

12 Arunachal Pradesh 2.7 16.0 37.7

13 Nagaland 4.8 14.7 43.6

14 Manipur 28.5 43.6 46.4

15 Mizoram 18.6 38.6 33.2

16 Tripura 19.3 12.6 25.9

17 Meghalaya 12.6 14.1 36.5

18 Assam 27.6 21.2 31.5


49

19 West Bengal 25.1 27.1 29.2

20 Jharkhand 9.9 14.2 25.0

21 Odisha 23.1 19.4 26.8

22 Chhattisgarh 13.3 28.5 22.6

23 Madhya Pradesh 12.9 30.4 31.3

24 Gujarat 20.7 13.4 27.4

25 Daman & Diu 13.9 7.3 7.9

26 D & N Haveli 8.2 7.4 15.9

27 Maharashtra 15.9 21.3 24.5

28 Andhra Pradesh (undivided)

27.3 25.5 35.9

29 Karnataka 13.7 17.2 29.5

30 Goa 20.4 13.5 23.6

31 Lakshadweep 17.4 16.2 20.1

32 Kerala 18.0 34.8 24.2

33 Tamil Nadu 11.2 27.7 22.4

34 Puducherry 13.7 18.1 11.2

35 A & N islands 17.2 11.9 20.6

36 Telangana# 25.0 24.1 -

37 Andhra Pradesh# 29.1 26.7 -

All-India (Urban) 61.6 25.0 24.9#newly formed states of Telangana and Andhra Pradesh

*Note: Economic Census is yet to release the data on the state-wise number of workers in all types of establishments by sector, establishment structure, and gender. That data, once made available, would open up scope for a more accurate comparison and triangulation of findings from these two national surveys. Check figure A-4 for the difference in% ‘no-travel’ and% establishments ‘inside household’.


50

Appendix VIII (continued)

Figure A-4: Difference between ‘ percentage of no-travel recorded in the Census’ and ‘ percentage of workplace inside household recorded in the Economic Census’


51

Appendix IX Vehicle ownership in households

Table A-14: percentage of households possessing particular type of vehicle

Population Population Density Households possessing a vehicle

Bicycle MTW Car

> 80 Lakh Very High 48.9 22.0 8.2

High 31.2 30.7 15.7

40-80 Lakh Very High 38.3 39.7 11.1

High 42.1 30.4 11.2

20-40 Lakh Very High 41.7 37.0 17.6

High 32.9 38.9 11.8

Medium 41.5 41.9 10.6

Low 61.6 42.3 11.7

10-20 Lakh Very High 18.1 36.6 13.7

High 40.3 38.0 9.6

Medium 41.4 35.2 8.8

Low 58.6 36.4 7.0

5-10 Lakh Very High 56.1 31.9 12.4

High 39.3 31.7 8.1

Medium 42.0 36.1 7.7

Low NA 36.7 8.4

1-5 Lakh Very High 70.0 20.7 3.6

High NA 27.5 6.8

Medium NA 31.1 7.2

Low 37.7 35.5 7.4

< 1 Lakh Very High 52.1 14.7 7.7

High 50.1 18.6 3.5

Medium 48.3 26.9 9.5

Low 45.8 NA NA

Transportation research and Injury Prevention Programme (TRIPP)

Indian Institute of Technology Delhihttp://tripp.iitd.ernet.in

Documents

TRAVEL TO WORK IN INDIAtripp.iitd.ac.in/assets/publication/WorkTravelReport.pdfTransportation Research and Injury Prevention Programme, the Indian Institute of Technology Delhi. This