Purchase decision of apartments in metropolitan India

Purchase decision of apartments in metropolitan India Factors affecting the purchase decision of apartments in metropolitan India Abstract Purpose The purpose of this paper is to provide an insight into the motivation behind Indian buyers when looking to purchase an apartment. The factors driving demand preferences for apartments are not well established and are difficult to measure, and often builders may not have an insight into what buyers are looking for. Design/methodology/approach The research in this paper is based on telephonic interviews and internet based survey with recent purchasers, who bought a home in the past 1 year and prospective purchasers looking to buy an apartment in the coming one year. They belonged to number of locations across all metropolitan cities of India Delhi, Mumbai, Bangalore, Kolkata and Chennai. The data were analysed using factor analysis to identify the criteria in an apartment that buyers value the most. This research was done across all ages and irrespective of their intention of why they bought or if this was their first purchase. Further, Cluster analyses was used to determine clusters and one way Anova was used to determine the factors that hold different value to different clusters of people. Discriminant Analysis was used to determine any difference in behaviour of first time purchasers with others. Findings The findings in this paper revealed that issues signifying â€Å"affluence† accounted for approximately 27 percent of the choice of housing by Indian buyers to purchase apartments in metropolitan India. Also, Cluster Analysis revealed that demographically different set of buyers differ significantly in their attitude towards â€Å"Financial† factors. Discriminant analysis revealed that first time buyers give significantly more importance to â€Å"Financial† factors like â€Å"House price†, â€Å"Income† where they give much lesser importance to â€Å"Builder reputation† and â€Å"Status of neighbourhood†. Research limitations/implications The research in this paper is aimed specifically at Indians living in metropolitan cities only which may be very different from the rest of India. The majority of the respondents belong to Delhi, which may also bias the results. The majority of the data has been collected from an online survey which may reduce the validity of the findings. Practical implications If due consideration is given to the factors that buyers are most concerned about, builders of new apartment housing would be better equipped to meet this demand and maximise their profits. Builders will also be able to target buyers better by knowing the difference in preference of first time buyers to others. Originality/value This paper provides an invaluable insight into Indians concept of a suitable apartment in metropolitans. While important decision factors were determined for the entire population, further analysis was done to determine difference in issues felt important to first time buyers. Also, the most important factors were determined for different demographic clusters. Thus in this way, the transaction of purchasing an apartment was analyzed from several points of view. Keywords Consumer behaviour, Purchase, Apartment, India Paper type Research paper INTRODUCTION The Real Estate sector is important to the Indian economy. In terms of employment generation, it is second only to the agricultural sector. The housing sector contributes nearly 5% to Indias GDP. It is expected to rise to 6 per cent in the next five years. Property markets in India are recovering faster than those in the US and the UK. The sector is expected to attract around US$ 12.11 billion of investments in the next five years. Residential space comprises almost 80% of the real estate developed in the country. There is a shortage of 22.4 million dwelling units according to the Tenth Five Year Plan. 80 to 90 million housing units will have to be constructed over the next 10 to 15 years to rectify this, with the majority of them for the middle- and lower-income groups. It is for this reason that residential properties in India, particularly in Mumbai and Delhi, are viewed as very good investments as per a study by PricewaterhouseCoopers (PwC) and Urban Land Institute, a global non-profit education and research institute. In the 2009-10 budget, a tax holiday on profits was granted to developers of affordable housing (units of 1,000-1,500 sq ft). This exemption was instituted for projects that started from 2007-08 onwards with a deadline of completion of March 1, 2012. US$ 207 million was also allocated to grant a 1% interest subsidy on home loans up to US$ 20,691 with the caveat that the cost of the home should not be more than US$ 41,382. This was expected to further help the housing sector. An apartment is a residential unit that forms a division of a building. It can be either owned or rented. Some people own their apartments together where each owns a part of the corporation which owns the flat. In condominiums, dwellers own the individual apartments and share the public environment. Living in apartments is gaining popularity in India. 217 townships across India are in the building plans for the Sahara Group. Their allure lies in the convenience that they offer in terms of safety and security and maintenance of utilities like electricity and water. A central maintenance system obviates the need for hiring outside help for minor problems like leaking taps or electric short circuits. Stand-alone homes also require incurring additional costs like buying/leasing land, licensing, duties, etc. Apartments enable maximization of space utilization and reduce demand on public resources. People are also able to avail of additional amenities like gymnasiums, swimming pools, etc. at affordable prices. There is a gap in the literature, however, with regard to the value drivers that dictate purchase decisions of residential property in the country. Similar studies exist for other countries but were found wanting in the Indian context, especially when it comes to apartments. Through this paper, we aim to do the very same, i.e. establish which factors dictate purchase decision and to what extent. We will also correlate these preferences with the demographic profiles and characteristics of our respondents and hence arrive at a greater and much deeper understanding of these issues. We see immense utility for our paper, especially for builders and property dealers who can use our findings in structuring their own business activities. RESEARCH BACKGROUND AND HYPOTHESIS Even though consumer behaviour is generally assumed to be an important part of real estate valuation, buyer preferences are generally not considered during the valuation process. It is basically reduced to the confirmation of a bid price which may or may not be met by the buyer. Efforts are being made to address this fault and many papers have been written on the analysis of motivations of residential property purchasers, attempting to explain them using models such as bounded rationality and hedonic pricing. Hedonic Pricing, or Hedonic Demand Theory as it is also known, decomposes the item of interest into constituents and evaluates the importance of each of them and their contribution to the overall valuation. These factors can be both internal characteristics of the good or service and external factors. In the case of real estate valuation, internal characteristics include layout, structure, etc of the property while status of neighbourhood, proximity to schools, etc are the exter nal factors. Factor Analysis enables us to do just that. It is a statistical method that reduces the number of variables by grouping two or more of them into unknown or hidden variables known as factors. Further analysis is then conducted by looking at the variation among these factors and evaluating their relative performance. These factors are taken to be linear combinations of the original variables plus error terms (Richard L. Gorsuch, 1983). â€Å"Factor analysis seeks to do precisely what humans have been engaged in doing throughout history that is to make order of the apparent chaos of the environment† (Child, 1990). It has great use in evaluating consumer behaviour. Charles Spearman is credited with its invention. He used it in the formulation of the ‘g Theory as part of his research on human intelligence (Williams, Zimmerman, Zumbo Ross, 2003). Over the years it has found uses in fields as diverse as psychometrics, marketing, physical sciences and economics. It can be used to segment consumers on the basis of what benefits they want from the product/service (Minhas Jacobs, 1996). It has evolved as a technique over the years, with many researchers working on fine-tuning and improving the analytical process. Bai Ng (2002) developed an econometric theory for factor models of large dimensions. It focused on the determination of the number of factors that should be included in the model. The basic premise of the authors was that a large number of variables can be modeled by a small number of reference variables. Marketing strategies based on customer preferences and behaviour often make use of this technique during the market research phase (Ali, Kapoor Moorthy, 2010) and while devising and changing the marketing mix (Ivy, 2008). Factor Analysis has also been used in ground water management to relate spatial distribution of various chemical parameters to different sources (Love, Hallbauer, Amos Hranova, 2004). The facility of segmentation that factor analysis offers has been extended to the real estate sector and all studies thereof. Regression analyses are subject to aggregation biases and segmented market models yield better results. This segmentation is done using factor analysis Watkins, 1999). Property researchers have also dedicated a lot of attention to researching the preferences of property buyers and identifying the drivers of property value. A study in Melbourne, Australia (Reid Mills, 2004) analyzed the purchase decisions of first time buyers and tried to determine the most influential attributes that affect the purchase decision using factor analysis. The research findings of the paper indicated that financial issues explain about 30% of the variance in the purchase decisions of first time house-owners. This related to timing, the choice of housing, and the decision to buy new housing. Apart from that the choice of housing is dependent on Site Specific factors (Location) and the decision to buy new housing is dependent on Lifecycle factors, such as family formation, marital status or the size of the existing house. Another study determined that brand, beauty and utility play a defining role in property value (Roulac, 2007). The findings of the paper explain why certain properties command premium prices, relative to other properties. It came to the conclusion that for value determination of high priced properties the overall perception of the brand is the most important factor followed by utility and beauty. Brand names are also very important especially in metropolitan markets as they add to the appeal, distinctiveness of the property. Another way to attract buyers attention is through the mix of neighborhood amenities offered (Benefield, 2009). Neighborhood amenities like tennis courts, clubhouses, golf courses, swimming pool, play park and boating facilities significantly impact property values. Xu (2008) used a hedonic pricing model to study the hous ing market of Shenzhen, China. He operated under the assumption that buyers consider property specifics and location attributes separately when they buy a home. The findings suggest that the marginal prices of attributes are not constant. Instead, they vary with the household profile and location. Cluster analysis involves the grouping of similar objects into distinct, mutually exclusive subsets known as clusters. The objective is to group either the data units or the variables into clusters such that the elements within a cluster have a high degree of natural association among themselves while the clusters remain relatively distinct from one another. Mulvey and Crowder (1979) presented and tested an effective optimization algorithm for clustering homogenous data. Punj and Stewart (1983) reviewed the applications of cluster analysis to marketing problems. They presented alternative methods of cluster analysis to evaluate their performance characteristics. They also discussed the issues and problems related to use and validation of cluster analysis methods. Ketchen and Shook (1996) chronicled the application of cluster analysis in strategic management research. They analyzed 45 published strategy studies and offered suggestions for improving the application of cluster analysis in future inquiries. They believed that cluster analysis is a useful tool but the technique must be applied prudently in order to ensure the validity of the insights it provides. Since Marketing researchers were introduced to discriminant analysis half a century ago, it has become a widely used analytical tool since they are frequently concerned with the nature and strength of the relationship between group memberships. It is especially useful in profiling characteristics of groups that are the most dominant in terms of discrimination. Morrison (1969) explained how discriminant analysis should be conducted using canned applications and how the effect of independent variables should be determined. However, care must be taken when applying discriminant analysis. The potential for bias in discriminant analysis has long been realized in marketing literature. Frank, Massy and Morrison (1965) showed that sample estimates of predictive power in n-way discriminant analysis are likely to be subject to an upward bias. This bias happens because the discriminant analysis technique tends to fit the sample data in ways that are systematically better than would be expected by chance. Crask and Perreault (1977) looked at the validation problems in small-sample discriminant analysis. Various research papers have studied the features that are evaluated while purchasing a home, how these features factor in terms of pricing the residences and how the home owners rate the various scales on importance. Such studies, however, were found lacking in the Indian context. This paper aims to understand the value drivers of apartments in Indian metros using factor analysis. The initial variables that we have considered are as follows Ø House Price This refers to the price/rent that is being charged for the apartment. The real estate market is often segmented using this variable. Ø Availability of Gymnasium, Swimming Pool and other sports facilities Many apartment complexes and housing societies offer recreational facilities to the residents to service their lifestyle needs. Ø Traffic This variable refers to the density of vehicular movement in the location in which the apartment is located. Ø Size of Individual Rooms The size of the rooms within the apartment is also an important factor. Some buyers prefer big, airy rooms while others might want smaller rooms. Ø Proximity to City This refers to the location of the apartment relative to the city boundaries, i.e. whether it is within the city proper or on the outskirts. Ø Ability to obtain Loans This variable stands for the ease with which the buyers can get loans, either through the builder or on their own. Ø Parking Space The availability of parking space is considered important by some consumers. Ø Exterior Look of the Apartment This refers to the faà §ade of the apartment, i.e. whether its attractiveness is a strong enough motivation. Ø Household Income The total income of the household often dictates the purchase decision of families. Ø Perceived Safety of Locality This is a big concern for some customers, especially single women and old people and may significantly influence the purchase decision. Ø Branded Building Components Some consumers may value an apartment more if it has branded fittings, furnishings, etc. Ø View from the apartment This can be an important variable for some customers. Ø Preference for Ground Floor This variable refers to the customers preference for the ground floor relative to other floors. Ø Water Supply This variable means to measure how important it is for the consumers that there is continuous, guaranteed and good quality water supply. Ø Structure This refers to the layout of the apartment whether it is a 2BHK or 3BHK, etc. Ø Status of Neighbourhood For some consumers, the reputation and social standing of the locality that they live is very important. Ø Proximity to Shops and Parks This seeks to measure whether proximity to these places is an important criterion for buyers or not. Ø Interior Design This refers to interior features of the apartment like flooring, lighting, balcony, etc. Ø Availability of Domestic Help This can be important consideration, especially for working couples. Ø Proximity to Schools and Offices This seeks to ask how important such proximity is to the buyer. Ø Builder Reputation Many buyers are heavily influenced by the brand name and reputation of the builder. Ø Monthly Living Costs Certain average monthly expenditure is incurred as living expenses. We seek to gauge the relative value of this variable. Ø Proximity to Public Transport, Major Roads, etc This refers to the accessibility of the apartment with regard to public transport and roads. Ø Power Backup Full power backup in case of power outages is frequently advertised by builders. Whether this actually influences buying behavior needs to be examined. Ø Proximity to friends/relatives homes This can be a big variable that dictates consumers in their decision-making process. Methods Sample The questionnaire was sent to people residing in Indian metropolitan cities. Out of the 172 responses received, 13 were rejected since the respondents had not purchased a property in a metropolitan city. Another 13 were rejected because either the respondents had not purchased the apartment in the last one year or were undecided as to when to purchase the property. Finally out of all the respondents 146 (84.9%) were identified. Measures The 25 variables were measured by a Likert scale with responses ranging from 1 (Very Low Importance) to 5 (Very High Importance). Analysis This study uses four tests to analyze the factors involved in purchase of an apartment. The first test conducted is the factor analysis which is used to club the variables in order to determine the purchase criteria of apartments. Thus, in this analysis the broad set of variables will be constricted to determine the smaller set of factors that can explain what home owners look for when purchasing an apartment. After this, a cluster analysis was conducted to determine the various clusters (groups) that exist within the demographic population. On the above said factor analysis and cluster analysis, a one way ANOVA was conducted in order to determine the order of preferences of each factors amongst such clusters. Finally, a discriminant analysis was conducted to identify factors that best differentiate the first time purchasers with others. Results The first test conducted was the factor analysis. Under this test, we followed the Principal Component Analysis method on the 25 variables to combine the correlated variables into factors. The KMO value calculated is 0.799 is above the suggested value of 0.5 which indicates that it is good idea to proceed with Factor Analysis. On the basis of the computations as represented in the Rotated Component Matrix (Table 1), the following factors were received: Affluence, Financial, location, lifestyle, Site-Specific. The variables were classified into a factor if their loading for the respective factor was greater than 0.4. Also, two other unnamed factors were received which remained so due to the fact that no factor can be formed between two variables. We have followed the Kaiser criterion (1960) of retaining only those factors that are greater than one. The initial research on 25 variables was reduced as the variables on domestic help, floor and proximity to friends/relatives was removed a fter the factor analysis was done. Domestic help was removed because it loaded on three factors (Financial, Location and Lifestyle) equally. Preference of Ground Floor was removed from the analysis as it showed a positive loading and negative loading on each of two factors which means that while some considered ground floor to be in consideration other considered the penthouse to be better. Proximity to friends/relatives was removed as it was the only variable in factor 6 (unnamed) and thus no factor can be made by one variable. The results of the Factor Analysis are as under: Rotated Component Matrix Variable Name Affluence Financial Location Lifestyle Site-Specific Unnamed Unnamed Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Factor 6 Factor 7 Traffic 0.768 Gym/Pool/Sports Facility 0.755 View from Apartment 0.721 Builder Reputation 0.644 Parking Space 0.568 Status 0.513 Monthly Cost of Living 0.764 Household Income 0.735 Availability of Loan 0.691 Availability of Domestic Help 0.498 0.414 0.435 Proximity to Schools/Office 0.778 Proximity to Transport 0.607 Proximity to City 0.575 0.424 -0.401 Proximity to Shops/Parks 0.546 Interior Design 0.768 Branded Components 0.712 Power Backup 0.594 Structure 0.741 Size 0.580 0.598 Safety 0.549 Preference of Ground Floor -0.415 0.423 Proximity to Friends/Relatives 0.845 Water Supply 0.410 0.652 House Price 0.405 0.508 Exterior Look 0.426 0.405 -0.464 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. Rotation converged in 21 iterations. Table 1 Factor Loadings- Purchase of an Apartment Table 2 Factor Analysis Factor No. Factor Name Eigen Values Total Variance (%) Cumulative Variance (%) 1 Affluence 6.826 27.306 27.306 2 Financial 2.9 11.600 38.906 3 Location 1.835 7.342 46.248 4 Lifestyle 1.504 6.016 52.264 5 Site-Specific 1.447 5.788 58.052 6 1.129 4.516 62.568 7 1.059 4.236 66.804 The second test that was conducted was the Cluster analysis and has done to segment the respondents on demographic variables of Age, Gender, City and Number of members in the family. Squared Euclidean distance and average linkage hierarchical clustering method was used. At fusion coefficient value of 1.0, two distinct clusters were evident. On conducting a One way ANOVA to compare means with the demographic variables we observe that the two clusters are differ on the mean age with a significance of 0%. The first cluster consists of a younger population with an average age of 37 approximately and the s