Socio-Economic Factors influencing Car Ownership by Indian Households

Rajesh Shukla, Anuj Das and Charu Jain (NCAER, New Delhi)

The automobile industry has witnessed a healthy growth over the last two decades, presenting huge opportunities for the automobile companies, both in two-wheeler segment and passenger car segment. Seeing the current scenario, it is utmost important for the marketer to understand the trends and happenings in the automobile market. National Council of Applied Economic Research (NCAER) conducted a survey¹ with the main objective of collecting information on income & expenditure, ownership pattern, consumer behaviour and entire socio-economic aspects of Indian households. As a supplement to this survey, another survey was conducted, which throws light on car market, named as "Demand for Car Market".

Taking the data of this survey as a baseline, this paper has tried to identify a number of factors that need to be analysed before segmenting the car market. Like, the marketer would like to know if there is any relationship between the socio economic & demographics factors of households with their purchasing power. How does the living style, educational level, source of income, durable ownership pattern influence the car ownership likelihood of a household? Is it the income or expenditure pattern, which should be considered most for capturing the right kind of market? The authors have attempted to answer such questions through the technique of probit regression analysis.

Introduction

The Indian automotive industry has flourished like never before in the recent years. On the canvas of the Indian Economy, Auto Industry occupies a prominent place. Due to its deep forward and backward linkages with several key segments of the economy, automotive industry has a strong multiplier effect and is capable of being the driver of economic growth. A reasonably developed Indian automotive industry ably fulfils this catalytic role.

Although the automotive industry in India is nearly six decades old, it remained dormant until 1982, largely due to stifling licensing regime. Today, India is the world's second largest manufacturer of two wheelers, fifth largest manufacturer of commercial vehicles and manufactures largest number of tractors in the world. The country offers fourth largest passenger car market in Asia.

During the year 2005-06, the turnover of the automotive sector was around $ 30 billion. According to auto industry experts, Indian Automobile sales will grow at a CAGR of 9.5% to 13,008 million units by 2010 from the current 10.0 million units.

This extraordinary growth that the Indian automotive industry has witnessed is a result of a two major factors namely, the improvement in the living standards of the middle class and an increase in their disposable incomes. Moreover, the liberalization steps, such as, relaxation of the foreign exchange and equity regulations, reduction of tariffs on imports, and the banking reforms, initiated by the Government of India, have played an equally important role in enabling the Indian Automotive industry achieve great heights. Also, the institutionalisation of automobile finance has further paved the way for a sustainable long-term high growth of the industry.

The once highly protected Indian automobile has been gradually opened up to the global market with liberalisation of overall economy. Global auto giants are shifting their manufacturing bases to India. Currently, almost all auto giants from Korea, Japan, US and Europe are present in Indian market in various segments. Foreign Direct Investment (FDI) has created a strong visible impact on the Indian car market not only by contributing to capital, technology and best managerial practices, but also introducing intense competition among the manufacturers.

The overall automobile industry performance has showed encouraging results for all the segments of the industry. Today, India has become the second fastest growing car market in the world. Passenger car sales have tripled in six years. It's also to be noted that the demand for luxurious models, SUVs, and mini-cars for family owners have shot up, largely due to increase in the consumer's buying capacity.

Clearly, the Indian automobile sector is on a growth track. To tap this huge opportunity, Indian automobile companies and global automotive giants are on an expansion mode. What is making the Indian automobile market grow? This paper is an attempt to answer such questions, which may help the marketers to segment their market effectively. We believe that our analysis and outlook of Indian automobile industry would serve as a key input for the business decisions and segmentation of Indian market for future demand. The paper focuses on the urban passenger car market and involves the critical analysis of car user households based on the following:

Studying the socio-economic and demographic characteristics of urban car owning households;

Understanding the relationship between various household characteristics and car purchasing intentions of households; and

Designing the likelihood model for car ownership by Indian households.

In the current study, a number of literatures and papers on the subject have been reviewed, which revealed that there are only few models suitable for such kind of analysis, which involves dummy regressands specially binary or dichotomous. These models are also known as probability models. Keeping in mind the objective of the study and availability of the kind of data or variables for conducting this analysis, the probit regression analysis is applied.

Data Source and Methodology

Data Source Used

For the current study, the data collected by National Council of Applied Economic Research (NCAER), in its annual survey "National Survey on Household Income and Expenditure 2004-05" is used. NCAER is conducting this survey every year since 1985. The survey aims to collect information on socio-economic aspects of the household viz. demography, employment, income, consumption expenditure, ownership pattern, etc. It covers 206 million households at all India level. Along with the main survey, another survey is undertaken with the objective of understanding the consumer behaviour of Indian car market by studying the purchases and demand patterns of Indian households. This survey, inter alia, also collects the primary data on car and two wheeler owners at national level.

General Characteristics of Indian Households/Families

It has been estimated that in 2005, there were about 206 million households in India, of which 61 million (30 per cent) resides in urban. Of the total households, nearly 9.4 million (5 per cent) households own at least one car, while 32 per cent household own two wheelers (and not cars). Although, just 30 per cent households are in urban areas but about 10 per cent of them own a car, while this ratio is just 2 per cent in case of rural. That is why the paper focuses on urban car market because this is the place where maximum growth in car demand is expected in near future.

Table below gives a clear comparative picture of demographic profiles of Indian households. It divides households at all India level into three categories: car owners, two wheeler owners (not having a car) and households with no automobile. It shows that the average annual income of a household in case of car owners is Rs. 1.99 lakh, which is much higher than the household owning two wheelers (Rs. 83,184).

About three-fourths of the car owning households are salary earners and self employed (non agriculture), while this is just little above 50 per cent in case of two wheeler owning households. Though both of these sources of income have equal shares in case of car owning households, there exists a huge difference within two wheeler owning households.

Table: Demographic Profile of Indian Households

	Car owners	2-wheeler owners not having car	HHD with no automobile	Total HHD
Share of households (%)	5	32	63	100
Average household income (Rs./annum)	198,966	83,184	40,457	62,518
Distribution of households by major source of household income %
Regular salary/wages	38	33	10	18
Self employed (non-agriculture)	37	24	12	17
Self employed (agriculture)	17	29	30	29
Labour	1	3	17	12
Other (Rent, pension, bonds, etc.)	8	11	32	24
Total	100	100	100	100
Distribution of households by highest literacy %
Illeterate	0	1	7	4
Upto higher secondary	31	60	82	72
Graduation +	66	37	10	22
Others	3	2	2	2
Total	100	100	100	100

Source: NCAER

 
Among those who have neither of the vehicles, just 10 per cent are salary earners.

The most interesting point to note here is that among car owning households, no household is found illiterate (i.e. at least one member in the household is educated). In fact, 66 per cent of the car owning households have at least one member with graduation or post-graduate education. In case of two wheeler owners, majority of the households (60 per cent) are educated upto higher secondary. It shows that in India, car owning households are comparatively well educated than others and have better occupational profiles.

Due to the fact that, of the total 9.5 million car owning households at all India level, 6 million are in urban areas, our discussion is hereafter restricted to urban India only. The survey result shows the following characteristics of urban car owners:

▪	Nearly 65 per cent of car owners are in urban areas.
▪	Around 90 per cent of them belong to higher income classes (fourth and fifth quintiles).
▪	The average annual household income of car owning households is Rs. 206,556 and average expenditure is Rs. 108,664.
▪	Though their average household size is 5 but average number of earners is just 1.4, which shows that the earlier saying in Indian families "More members in family adds to more household income" is no more valid.
▪	Zonal distribution of car owning households shows that majority of them are in south (35 per cent) followed by west (30 per cent). 23 per cent of them reside in north.
▪	About 66 per cent of car owning households also own a two-wheeler.
▪	The car purchase decision in Indian urban families is influenced by the head of the household followed by the chief earner. Female members contribute just 14 per cent in making such decisions.
▪	Only 29 per cent of car urban car owning households have unmarried daughter (above 16 yrs of age), and around 80 per cent of these households are free from students liability for higher education.
▪	4 per cent of such households possess land ownership above 4 acres.
▪	The durable ownership patterns are also very impressive in car owning households. 93 per cent of them own colour TV, refrigerator (90 per cent), cellular phone (82 per cent), washing machine (66 per cent), motorcycle (51 per cent), scooter (21 per cent), and credit cards just 19 per cent.

 
The socio-economic structure of households is one of the important factors affecting the car market. With increasing purchasing power, the number of households crossing the threshold plays an important role in passenger car sales. Therefore, from our analysis point of view, we have used some of these socio-economic variables like total consumption, primary source of income, number of earners, land ownership, household size, education of chief earner, ownership of dwelling unit, and ownership of durable goods in our paper.

Methodology

As mentioned earlier, the objective of the paper is to analyse all those variables that throw light on the factors affecting the ownership of car by a household. To develop a model that could define the likelihood of owning a car, we have divided all the households in two categories: those who own a car, and those who do not own a car. It means that our regressand is a dummy variable.

Probit model

Keeping in view the type of analysis to be conducted, in current study, we have used pro bit regression model.

The main purpose of using this model is that unlike other models, this is not limited by the problems like not bounded {0, 1} and hetroskedastic errors. Probit regression improves results as they give predicted probabilities between 0 and 1 by 'smoothing out the tails'. The choice of the pro bit is done by observing the means of dependent variable. If it lies between .2 and .8, then probit is preferred and if the mean of the dependent is more extreme then logit is preferred, which is similar to the earlier one.

Our interest lies in identifying and studying the socio-economic and demographic factors that are affecting the car purchasing intention of a household. Any household's economic wellbeing is related to its income, consumption expenditure and major source of HHD income. It also depends upon the number of earners in a HHD, chief earner's education and land owned by HHD. Besides, durable goods like colour TV, refrigerator, motorcycle etc., and ownership of dwelling unit and its structure also play an important role in explaining the lifestyle of a HHD. Considering the Indian scenario, factors like children's education and marriage of daughters also playa significant role in durable products (e.g. car) purchase.

Thus, our model can be explained as:

 
Where 'Y' is the dependent variable and ‘I’ is the probit score or index. B’s are regression coefficients and X's are explanatory variables.

Here, dependent variable would be "Car ownership of a household" against which above discussed socio-economic and demographic factors would be tested. Thus, our dependent variable will take only two values -

1- If the concerned HHD owned a car and

0- Otherwise.

Probit regression was run with different socio-economic demographic variables in STATA, and finally came up with some 18 socio-economic and demographic factors, which were having the maximum impact on dependent.

Results Interpretation & Validations

Interpreting the coefficients of explanatory variables obtained from probit analysis is illogical. Thus, we will interpret them in terms of probability. On running a probit analysis, we found that all the variables are statistically significant at 5% level of significance, except number of unmarried daughters, fire insurance and theft insurance.

Holding all other variables constant, if the yearly total income of a HHD is going to increase by Rs. 1 lakh, then the probability that the household will own a car increases by 0.8% only. Similarly, the probability of car owning shows a positive increase of 19.2% with increase of Rs. 1 lakh in HHD's yearly consumption expenditure. This implies that in urban India, HHDs with better economic condition, which are spending more on their consumption expenditure has a higher chance of owning a car.

If we consider the HHD's primary source of income, then those HHDs whose income primarily comes from business has a higher probability (13.8%) of owning a car than HHDs with salary (8%) as the main source. It clearly reflects that business class families are more likely to own a car than salary ones. In urban India, HHDs whose main income comes from agriculture has a 16.7% probability of owning a car. This shows that agriculturists in urban India are generally richer than their rural counterparts. Also HHDs having income from other sources have a probability of 17.2% of owning a car.

Education of chief earner is showing a positive impact on car ownership. The probability of owning a car by a HHD is higher by 5.1 % when the chief earner's qualification is graduation or above as compared to HHDs with chief earner's education below graduation. It is interesting to see that number of earners is showing a negative impact on car ownership of a HHD. With increase in one unit, i.e. one earning member, the chance of owning a car decreases by 2.1 %. This shows that large number of earners in a HHD doesn't imply that its economic condition is strong. Similarly, with a unit increase in HHD size, the car owning probability of HHD decreases by a unit (1%), showing a negative impact on car ownership.

Though number of unmarried daughters is not significant at significance level of 5%, yet, it is showing a negative impact (0.7%) on car ownership probability. Interestingly, number of students pursuing higher education in a HHD is having a positive impact on car ownership. A 2.2% increase in car ownership probability of a HHD is observed, with a unit increase in the number of students in that HHD.

Land ownership is playing an important role in car ownership of a HHD. It is observed that a HHD having 4 acre and above land has a higher chance (4.3%) of having a car than those having land ownership below 4 acres. HHDs owning dwelling units with "Pucca" structure have a positive impact than HHDs owning dwelling units with "Semi-Pucca" structure, which has a negative impact. A 3.4% increase in car ownership probability is observed for HHD with owned dwelling unit and dwelling structure "Pucca", whereas 3.9% decrease in car ownership probability is observed for HHD with owned dwelling unit and dwelling structure "Semi-Pucca".

The ownership of white goods has a high impact on car ownership. If a HHD owns CTV, motorcycle, refrigerator, cellular phone and credit card, then the car owning probability of that HHD increases by 13.2%. This shows that HHDs in urban India owning at least the above mentioned consumer goods have a higher chance of having a car. It is also observed that HHDs having life insurance are showing a positive increase of 4.7% on the probability of car ownership, whereas HHDs having fire insurance or theft insurance are having no impact on car ownership. This shows that urban India still thinks life insurance as an important economic security measure rather than fire or theft insurance.

We have used sensitivity and specificity to plot Receiver Operating Characteristic (ROC) curve. The area under the ROC curve measures the accuracy of the model. An area of 1 represents a perfect model; an area of 0.5 represents a worthless model. A rough guide for classifying the accuracy of the model:

▪	0.90-1 = Excellent
▪	0.80-0.90 = Good
▪	0.70-0.80 = Fair
▪	0.60-0.70 = Poor
▪	0.50-0.60 = Fail

 
 
On plotting the ROC curve, the area under ROC came out to be 0.90 (refer Fig. 1). This shows that our model is perfectly classifying the HHDs on the basis of car ownership.

Conclusion

The automotive industry in India is now finding increased recognition worldwide. With an impressive growth rate in sales, it is spreading its arms in the global markets as well. Advent of major global players in Indian market and their interest in making India as a manufacturing hub has made Indian auto industry very competitive due to availability of ready market.

Due to intense competition and immense opportunities available in Indian market, it is imperative for auto marketers to understand the customers in all aspects in order to survive in the market. Seeing present scenario, there is a need to segment the market in the right direction, which requires a complete strategic planning.

We feel that identifying the demographic profiles, lifestyles and socio-economic characteristics of households helps a lot in understanding the present market so as to devise strategies for the future. This paper is an attempt in this direction. Though, directly, the study is not analyzing the demand of cars in urban market in future, indirectly it is telling us about what factors need to be considered to tap the right kind of potential customers for cars.

We have found very interesting results showing that there are certain household demographic characteristics, which, if studied properly, by marketers could help them a lot in targeting the potential car customers in future. Major factors are broadly discussed below:

▪	Results show that while income has great influence on demand, it is the expenditure pattern that really determines the likelihood of purchase of major consumer durables, including car, by a household. The level of annual average expenditure by a household has a strong relationship with car purchase. The increase in annual expenditure by Rs. One lakh reflects the increase in status of a household in terms of both financial and social aspects and hence increases the chances of a household to go for a car as it is still considered as a status symbol in Indian society. Therefore, marketers who are totally focussing on income for targeting and segmentation of their markets should include expenditure part as well in their strategies.
▪	Secondly, the major source of household income also plays an important role in creating possibility of car purchase by a household. This is what our result shows. It clearly shows that those households with business as major source of income have greater chance of purchasing a car in comparison to households with salary as major source of income.
▪	Lastly, the product ownership (white goods) of a household also, to a great extent, determines the car purchase likelihood. Those households with presently owning at least colour television, refrigerator, motorcycle, cellular phone and credit card have greater chances of purchasing a car in near future. The reason is that these households already own all the major white goods; so, their next priority could be a status item like car.

 
To conclude, there are huge opportunities lying ahead for Indian automotive industry. Right kind of planning with effective segmentation of market, could help the marketers reap huge benefits in future.