A Study Of Passenger Car Buying Behavior And Demand Based On Household Survey Data

It is30years since China reform and opening-up, Chinese resident living level have takenplace great changes in this period. As resident income increased significantly, consumer behaviorchanged and consumption structure upgraded. On the one hand, the proportion of basic livingexpenses, for example food, decreased gradually; On the other hand, number of large durableconsumer goods is increasing. Especially after China joined WTO, as a typical durable consumergoods, automobile start to entry to Chinese family, the volume of private passenger car increasedtremendously, and China step into passenger car popularization period. The volume of passengercar in China surpass the U.S. and become the largest passenger car market in the world in2009.After2009, Chinese passenger car market not only keep the No.1position in the world, but alsoreached19.4million in2012, which renew a record in the world. Take this information asbackground, this paper take car as the research object. Automobile products are a kind of typicaldifferentiation characteristics of consumer goods. For consumer goods with high differentiationcharacters, reliability and authenticity of research results will be doubtful if we ignore thedifferentiation character. This paper will try to analysis automobile consumption behavior anddemand by heterogeneity assumption. For consumption heterogeneity assumption study, normallyit will be based on space feature study, as discrete choice model keep developing and improving, itprovides powerful analysis tools.By reviewing relevant studies, and considering current Chinese family automobileconsumption status, this article comes up with its study objective: taking Chinese familyautomobile consumption as research objective, by conducting logit model, to find out Chineseconsumer car purchasing behavior. Furthermore, to estimate consumer automobile consumptiontrends from micro data level.The key content of this article is Chinese household’s automobile consumption, and this articlecovers many types of data, there are macro data and micro data (such as sampling survey data). Themicro data covers almost all brands and models in China market which ensure thecomprehensiveness of the analysis. Below is the key content of this article:Chapter1of this article describes the consumption theory, and pointed out the mainquestions and research method of this article based on reviewing current research progress.Chapter2mainly stated three questions. Firstly, this chapter make a detailed review about thepresent situation of the development of the whole car market, especially about the development history of passenger car market. The second, using the province level data, this chapter study theautomobile market competition situation based on the concentration index and the concept oflocation quotient, find the major automobile manufacturers’ competition situation, particllarly weanalyse the aggregate consumer behavior characteristics by using these result. Finally, we specifythe national passenger car consumer survey data of2012that this article used, the representation ofthe sample data statistical characteristic is better, it can reflect the overall passenger car market’ssituation.Chapter3Firstly, we introduce the discrete choice model, understand the model theory and themain difference, especially focus on the nested logit model. Secondly, under the heterogeneity, wedefine the consumption behavior process, the buying process of consumer is assumed to be firstlybetween domestic brands and foreign brands, and secondly between the different level of car, thelast to choose between different product brands. This behavioral assumption is the expectation, andwe will build the mode based on this expectation. In the following chapter, we will tese theassumption. Finally, we choice the product brand that can be introduced in the model. In thischapter, we prepare three aspects for the next quantitative study, the three aspects are theoreticalmodel, behavioral assumption and product brands.Chapter4Firstly, we carry on the qualitative research on the saloon car. From the micro levelperspective of product differentiation, based on the random utility theory, we establish three levelsof nested logit model. Which unify the consumer characteristics and product characteristics in adiscrete choice model, and find the influenced factors of each level. Based on these factors, find theimpact to choosing probability for family. Research results show that the product attribute, carengine power, speed, wheelbase and prices for the consumer has a major influence on the choice ofproduct brand, body size, whether to have children, seat number for the consumer’s choice of carlevel has important influence, and consumers’ sex, age, level of education and income for theconsumers to choose from domestic brands and foreign have a significant impact. Secondly, wecarry on the same qualitative research on the SUV, and get the relevant conclusion. Finally, throughcomparison and analysis of the marginal utility of model results, we get the further conclusion.Chapter5we compute consumer demand for car by a new method. the first, This chapterintroduces the concept of "second choice". When the questionnaire survey was conducted onconsumers, consumers need to answer the question, that is at the time of purchasing, what is themost seriously considering second choice models. Through this problem to collect the secondchoice models’ information. By using the second choice data, we can estimate the relevantparameters by semiparametric estirnation and GMM. Finally get the probability of consumerchoice model.Chapter6make a volume forecaast for the total passenger vehicle market. In this paper, we use VAR model, get the stability relationship between variables and new car sales volume, and getthe total volume forecast. After getting the total volume, we use the product selection probabilitywhich is obtained from the previous section to calculate the approximate product sales volume.From the present literature retrieval, under the premise of considering product heterogeneity,correlative papers are few that using the consumer survey data at present. In this paper, the maininnovations are as follows:1. In this paper, on the basis of large sample, under the premise of considering productheterogeneity，we research the Chinese family car consumer choice behavior. Through three layersof nested logit model to construct the consumer choice process of car and estimate the model.Reveals the mechanism of the process of household buying car, which provide the theoretical basisfor the enterprise strategy and government policy.2.The most of the domestic automotive research are based on homogeneity premise, underthis assumption, it is difficult to indepth study of Chinese family car consumption behavior; even ifthere are a few studies which research related issues under the assumption of heterogeneity usingmicro data, but their study usually has certain representativeness problems because of the smallsample size. And the sample size of this paper is relatively large and the regional distribution isrelatively more extensive, Through indepth study on consumers purchasing behavior, I get someimportant results and make some progress in this study.3.In the use of "second choice" to compute the substitutional relation between models,and establish a complete set of requirements computing framework from the micro level and arelatively accurate demand quantity of the product can be calculated. Thus providing a complete setof method system in the car to represent the heterogeneity of strong consumer durablessegmentation varieties demand forecast.4.We study provincial passenger car consumption pattern based on registration data, in thestudy we use the concentration index and the concept of location quotient, find differentconsumption characteristics in the different provinces and different manufacturers of passenger carsconsumption pattern characteristics.This reveals the advantages and disadvantages ofmanufacturers in different provinces, provide a theoretical basis for a regional strategy of themanufacturers.Of course, because of the huge amount of micro investigation data, especially involvingdemographic information every family surveyed, the product information etc. and other dimensionsaggregated together which lead to a huge amount of calculation, as well as the paper’s lengthlimitation which lead to that not all models or brands be involved in this article, so some productswhich has the highest market share in its segment and the most representative products areanalyzed in paper. So there are some limitations in this paper which remains to be excavated in the future.Finally, This paper based on cross-sectional data. In the follow-up studies, we can use the timeseries or panel data, which can get the result about the change of behavior and demand.