A typology of Housing Search Behaviour in the Owner-Occupier Sector

Download 13.27 Mb.
Size13.27 Mb.
1   2   3   4   5   6   7   8   9   ...   60

List of Figures

Fig 2.1 A spectral diagram of the relative realism of assumptions about human behaviour in different schools of economic thought 32

Fig. 3.1: Alternative views of housing choice 75

Fig. 3.2 A conceptualisation of the different housing search stages and the recursive feedback links between each stage 76

Fig. 3.3: Aspects of variation across stages of housing search 93

Fig 4.1: A diagram of the spectrum of housing search behaviour from Neoclassical to Old Behavioural 98

Fig 4.2: Decision-making, search and choice: Tu and Goldfinch, 1996 100

Fig. 4.3: Schematic diagram of decision making model: Smith et al (1979) 101

Fig. 4.4: An adaptation of three models of mobility: Speare, Goldstein and Frey, 1975 104

Fig. 4.5: The chain of search stage: Maclennan and Wood, 1982 106

Fig. 4.6: Housing choice processes: Maclennen ,1982 108

Fig. 4.7: Housing search and choice: Wong, 2002 111

Fig. 4.8: Housing market mobility, search and choice decisions: Marsh and Gibb, 2011 115

Fig. 4.9: A framework for understanding online housing search: Rae, 2015 118

Fig. 4.10: Role players and family decision process: Levy and Lee, 2004 119

Fig. 4.11: Simplified model of the housing search process and key variables 121

Fig. 4.12: Existing housing search models on a spectrum from NCE to OBE 124

Fig. 5.1: An example of the Likert item questions in the survey 151

Fig. 5.2: An example of a text based question in the survey 152

Fig. 5.3: An example of a binary question in the survey 152

Fig. 5.4: Map of all sold dwellings in 2010 and survey responses 161

Fig 5.5: A depiction of the combined methods: categorical principal components analyses and cluster analysis for the construction of a typology 168

Fig 5.6: Median house prices to median earnings ratio, 1997 to 2013 184

Fig.5.7: Map of the 13 Housing Market Areas (sub-markets) in Sheffield, according to the 2013 Strategic Housing Market Assessment and Building Footprints 185

Fig 6.1 Map of location of all dwelling sales in Sheffield, 2010 198

Fig. 6.2: Mean time between events in search for household types 237

Fig: 7.1 Number of clusters in Ward CA and difference between coefficients 300

Fig. 7.2: Pre-search awareness net agree scores by cluster 306

Fig. 7.3: Pre-search awareness net agree scores by cluster (excluding cluster D) 307

Fig. 7.4: Pre-search attitude towards moving home net agree scores by cluster 309

Fig. 7.5: Motivations in first considering moving, net agree scores by cluster 311

Fig. 7.6: Motivations 2 – importance of factors by cluster 312

Fig. 7.7: Aspirations – what did you want from your dwelling – net agree by cluster 315

Fig. 7.8: Search length and time pressures, net agree by cluster 317

Fig. 7.9: Search length, by stage for 25th, 50th and 75th percentiles of cluster members 318

Fig. 7.10: Importance of information sources by cluster 320

Fig. 7.11: Frequency of information use by cluster 321

Fig. 7.12: Experience of searching for a new dwelling and outcome 323

List of Tables

Table 2.1: Behavioural underpinnings of housing economics 57

Table 3.1 Decision to form a household, adapted from Ferrari et al (2010b) 78

Table 3.2: Decision to move, adapted from Ferrari et al. (2010b) 82

Table 3.3: Selection of a dwelling, adapted from Ferrari et al. (2010b) 84

Table 3.4: Housing location search, adapted from Ferrari et al (2010b) 88

Table 3.5: Search strategy, adapted from Ferrari et al (2010b) 91

Table 5.1 List of survey variables and data types: across stages, household and dwelling characteristics 149

Table 5.2: Location: population and survey response rate (by Housing Market Area) 163

Table 5.3: Month of exchange completion: population and survey response rate 164

Table 5.4: New or previously sold property: population and survey response rate 165

Table 5.5: Property type: population and survey response rate 166

Table 5.6: Variables to be used in the CATPCA 176

Table 6.1: Average sale prices by HMA, 2007 and 2012, and the change in price 201

Table 6.2: Average rent per month in each HMA 202

Table 6.3: Lower quartile house prices and incomes needed to afford the lower quartile price with the percentage of households with income in each area able to afford a lower quartile priced dwelling. 203

Table 6.4: Housing stock in Sheffield by tenure in 2007 206

Table 6.5: Number of completions by type and year 2007-12 207

Table 6.6: Dwelling completions by HMA and year 2007-12 208

Table 6.7: Period of previous move 210

Table 6.8: Period of previous move by household type (percentage in each household type) 211

Table 6.9: Tenure of previous dwelling 212

Table 6.10: Previous tenure by household type (percentage in each household type) 213

Table 6.11: Households’ previous dwelling location 215

Table 6.12: Before searching, to what extent would you have agreed with the statements (1, Strongly Agree to 5, Strongly Disagree) 216

Table 6.13: Location, size and type preference awareness 219

Table 6.14: Percentage of households, before considering moving to what extent would you have agreed with these statements 223

Table 6.15: Motivations for moving, to what extent do you agree with the following statements? 225

Table 6.16: Motivations for moving dwelling, percentage of household types agreeing with statements (percentage in each household type) 226

Table 6.17: Importance of factors in initial decision to start searching for a property 227

Table 6.18: From very important to unimportant: What did you want from your home when you first considered moving? 229

Table 6.19: Percentage of responses Very important or Important: What did you want from your home when you first considered moving? (percentage in each household type) 231

Table 6.20: Percentage of all households participating first in an event in each year 233

Table 6.21: Percentage of household types participating first in an event in each year (percentage in each household type) 234

Table 6.22: Time between events in search for all households (descriptive statistics and frequencies) 236

Table 6.23: To what extent do you agree with the following statements about time, length and search restrictions 239

Table 6.24: Percentage of households by type who agreed (Strongly agree or Agree) with the following statements (percentage in each household type) 240

Table 6.25: Number of properties visited in person 241

Table 6.26: Number of properties (in groups) visited in person by household type (percentage in each household type) 242

Table 6.27: The number of properties households placed an offer on 243

Table 6.28: Number of properties placed an offer on by household type (percentage in each household type) 244

Table: 6.29: When event X occurred had the home you hoped for altered from your hopes before the event? 246

Table: 6.30: Percentage of households who altered their hopes by household type (percentage in each household type) 247

Table 6.31: Importance of different information sources in deciding current home 250

Table 6.32: Percentage of households indicating each information source is important or very important (combined) by household type (percentage in each household type) 251

Table 6.33: Percentage of households indicating frequency of use of information source 254

Table 6.34: Percentage of households indicating they used the information source very often or often by household type (percentage in each household type) 255

Table 6.35: When did you most often use the source of information? 256

Table 6.36: When did you most often use the source of information? Throughout and Never answers removed 258

Table 6.37: To what extent do you agree that your new home accomplishes your goals? 261

Table 6.38: To what extent do you agree that your new home accomplishes your goals? Only households who considered each of the aspirations Very important, Important or Moderately important. 263

Table 6.39: How would you describe your experience for searching for your current home? To what extent do you agree with the following statements? 265

Table 7.1. Variables used in the CATPCA with % responses to variables 290

Table 7.2 CATPCA model summaries 293

Table 7.3 Number of cases in clusters in Cluster Analysis solutions with different numbers of cluster specified 294

Table 7.4 Round two model summaries 296

Table 7.5: CATPCA a) Pre-Search component key variable loadings 297

Table 7.6: CATPCA b) Motivation components key variable loadings 298

Table 7.7: CATPCA c) Search component key variable loadings 299

Table 7.8 Number of households in each cluster for three to seven cluster solutions 300

Table 7.9: Previous tenure by cluster 304

Table 7.10: Year of last move 305

Table 7.11: Location of previous dwelling 306

Table 7.12 Summary of Pre-Search attributes by Cluster 316

Table 7.13 Descriptive statistics by cluster for number of properties visited and placed offer on 322

Table 7.14 Summary of search attributes by Cluster 324

Table 7.15 Summary of household attributes by Cluster 325

Table 7.16 Typology description by variables and précis of key attributes 326


Austrian Economics AE

Behavioural Economics BE

Cluster Analysis CA

Categorical Principal Components Analysis CATPCA

Critical Realism CR

Department for Communities and Local Government DCLG

Housing Market Area HMA

Institutional Economics IE

Land Registry HMLR

Marxist Economics ME

Neoclassical Economics NCE

New Behavioural Economics NBE

New Institutional Economics NIE

Office of National Statistics ONS

Old Behavioural Economics OBE

Old Institutional Economics OIE

Principal Components Analysis PCA

Travel to Work Area TTWA

Chapter One: An Introduction to Housing Search, Behavioural Economics and Typologies

Key points from chapter one

  • The process of searching for, and buying a home is a significant and extraordinary process for most households that engage in it

  • However, housing search has largely been neglected as a research subject for long periods in the twenty and twenty first centuries by the major schools of economics

  • Little is currently known about the variety of household behaviours in the search process and there is a dearth of empirics about actual behaviour

  • The renewed political and academic interest in the behaviour of households in the housing market suggests that this is an opportune time for a research project into the search process

  • This study aims to explore the variation in housing search behaviours and to construct a typology of those behaviours

1.1 Introduction

The purchase of a dwelling to live in constitutes a major event for the many people that undertake it. The process they go through, of searching for and purchasing a dwelling, is often cited as one of the most stressful periods in life (Raviv et al., 1990). It is dissimilar to the processes in the majority of their routine purchases; in which a known product, with a known utility is purchased from a known retailer at a known price. The search for a dwelling to own and live in is, in contrast, a deliberative step towards purchasing a product that has not been consumed previously, in a market that changes rapidly, at a price that is likely to be unknown precisely. For many households it does not feel like a “small world” problem (see Savage, 1954; Berkeley and Humphreys, 1982; Marsh and Gibb, 2011) that they make routinely as part of their everyday experience. For the household, at the experiential level, the search process itself is a significant aspect of the overall transaction. Despite the complexity of finding a dwelling to live in, with its emotional impact and uncertainty, many households surmount these difficulties and manage to move dwelling. Yet, relatively little is known about the process they go through or the impact that their search process has on market dynamics (Maclennan, 1982; Munro, 1987, Dunning and Watkins, 2012).

Economics, in an attempt at simplification, frequently skips the search process in its analysis of housing markets, focussing solely on the outcomes of moves (McCarthy, 1982; Smith et al., 2006; Marsh and Gibb, 2011; Butler and Hamnett, 2012). Where economic research does model the process, it frequently homogenises households and their behaviour. Arguably, however, the justification for simplifying households’ search behaviours is weak; the housing search process does matter; it intervenes at a crucial stage of the potential movement from dwelling to dwelling, in some cases determining if, to where and to what, a household moves (Wolpert, 1965). It is therefore essential to understand the search process if we are to understand moving behaviour and the housing market (McCarthy, 1982).

Intuitively, the complexity of the search and decision making process can be recognised, and also that households’ experiences are unlikely to be homogeneous. Households, with different characteristics, move dwelling for different reasons, they also have different aspirations and different search processes. The life events and life course literature suggests that births, deaths, changing jobs, leaving the parental home and other significant events often precipitate a move of dwelling and those households constitute different combinations of socio-demographic groups (Rossi, 1955; Clark and Dieleman, 1996; Groot et al, 2011; Van Ham, 2012).

Households, with different motivations for purchasing a dwelling, are presented with multiple sources to gather information about the market and vacant dwellings (Dunning and Watkins, 2012). These information sources are rarely neutral, frequently facilitated by agents with vested interests (Levitt and Syverson, 2008) who may use different linguistic and emotive devices in different market conditions and in different locations (Pryce and Oates, 2009).

The extensive variation in household characteristics and the complexity of gathering information about vacancies combine to make the housing search process a highly complex and variable economic activity. Therefore, approaches to the analysis of housing markets need to consider the variation in housing search behaviours, likewise policy makers interested in intervening in the owner-occupier housing market need to understand the differences between search behaviours before they design policies to influence outcomes (Ferrari et al, 2011; Gibb, 2013).

Interest in the functioning of housing markets and the economic activity of agents is understandably high given the impact of housing finance products on the wider economy through the Global Financial Crisis and subsequent fall in house prices in real terms across many housing markets (e.g. Smith and Searle, 2010). Neoclassical economic models, based on the rational behaviour of individuals, have however persistently failed to consider the observed behaviour of actors in markets (Simon, 1994; Marsh and Gibb, 2011) and subsequently have failed to predict market outcomes (Meen, 2003; Munro and Smith, 2009).

Despite the absence of housing search processes in most neoclassical economic accounts there have been intermittent calls to describe and explain the manner of searching for a dwelling and to recognise the importance of search as part of the overall dwelling purchase process (McCarthy, 1980; Clark, 1982; Maclennan, 2012). Greater attention is now being paid to the microeconomic foundations of housing than was the case in the latter half of the twentieth century (Watkins, 2009). The microstructures approach aims to better understand the behaviours of households in the housing market, and is now more frequently addressed through multidisciplinary alliances, in the hope of shedding more light on the causes of economic variability than neoclassical economics has supplied (Smith and Munro, 2009; Smith, 2013). Despite this need research into housing search behaviour has developed only spasmodically (McPeake, 1998).

Motivation for studying housing search behaviour

Outcomes act as signals to both buyers and vendors in the market place and to those seeking to shape or respond to market pressures, not least local authorities with responsibility for overseeing the development process and with political concerns about the state of the housing market. However, focussing only on outcomes precludes information that could lead to a more detailed understanding of those outcomes and therefore potential policy actions and interventions. Maclennan and O’Sullivan (2012) argue that studying search behaviour has the potential to reveal information about constraints, substitutability and latent demand that may not be easily observed from an outcome-only perspective, and is therefore of intrinsic interest to policy makers as well as housing researchers.

Housing economics has a history of focussing on the outcomes rather than the processes of housing search (Maclennan, 2012). But this is not the case in all fields of economics. Labour market search, for example, has a long history of relaxing the assumptions of perfect rationality and substituting satisficing behaviour in the job search process in order to explain spatial variation in the labour market (Wolpert, 1964; Clark and Moore, 1982).

The complexity and high costs of obtaining accurate information about buyer search behaviour is another reason why housing search studies have been less popular, but with the availability of new data sources on search behaviour this may change (Rae, 2014). Search behaviour is also being considered by organisations, such as the Bank of England, as early indicators of price changes and pressures within a market that may not have translated into increased sales volumes or prices yet (McLaren and Shanbhogue, 2011; Hohenstatt and Kaesbauer, 2014). While neoclassical accounts have marginalised search, not all economic approaches have ignored behaviour in the housing market.

The rationales for modelling housing search largely fall into two categories: enabling prediction of behaviour and/or outcomes (the normative); and, for clarification of conceptual differences between approaches to understanding the reality of the housing search process (the descriptive). The first rationale has been more popular and traditionally received more attention from economists (Simon, 1959), but is coming under more pressure as contestation over the predictive power of mainstream models increases (e.g. Marsh and Gibb, 2011a; but see Boelhouwer, 2011 for a rejection of this argument). Pursuit of the second rationale is also useful in order to provide a clear differential of the various models of housing search.

Justification for interest in Behavioural Economics

Interest in housing search and household behaviour has not arisen in isolation from wider trends in housing studies and land and property economics. One such driver is the recent revival in interest amongst both academic and policy makers in the potential use of ideas from behavioural economics to shape public housing policy (Hincks et al., 2013; Adams and Watkins, 2014). Driven by frustrations with the limitations of neoclassical economic models, and their weak behavioural and psychological underpinnings, attention has turned to alternatives in the work of behavioural economists such as Herbert Simon, Daniel Kahneman and Amos Tversky. Political attention is illustrated at a general level by the coalition government’s expansion of the Behavioural Insights Team and specifically in housing policy through DCLG’s commissioning of behavioural research (e.g. Ferrari, 2011a; Dunning et al., 2014a; Dunning et al., 2014b; Watkins et al., 2014). Evidence of the resurgence of academic interest is widespread at both the general economics level (Wilkinson and Klaes, 2012) and in the case of housing economics specifically (see below).

The revival of research interest in housing search has been heralded previously (e.g. Maclennan, 1982) but has been propounded by the focus issue of Housing, Theory and Society (Vol. 28, No.3, 2011). The issue is devoted to the topic of behavioural economics and housing theory, frequently turning to the issue of housing search. A short overview of the issues follows. Marsh and Gibb (2011) lament the failure of theoretical and empirical studies to build on earlier behavioural research that critiqued mainstream economics (e.g. Maclennan, 1982) and call for alternative, behavioural theorisations of search. Smith (2011) is a strong proponent of utilising behavioural insights in housing economics, particularly where augmented from sociological studies. Clark (2011), critiquing Marsh and Gibb’s and Smith’s papers, supports the recommendation for a further exploration of housing search and draws the links between housing search models and previous alternatives to the expected utility model (e.g. Clark and Smith, 1985). Boelhouwer (2011) remains sceptical about the failure of mainstream economics, but argues that where behavioural models are useful for understanding housing market processes, they also need to be useful for modelling purposes. The creation of search typologies fits Boelhouwer’s argument that there is a balance between detailed insight and abstractions to describe broader processes. Watkins and McMaster (2011) are supportive of the application of behavioural approaches to housing: “we argue that the failure of mainstream economics to adequately explain housing choice processes is central to the weakness of prominent econometric models” (Watkins and McMaster, 2011, P.281). Their support is caveated by the call for clarity in the theorisation of housing economics, warning against the assumption that behavioural approaches are uniform and mutually inclusive. Clapham’s (2011) critique is the most scathing, but largely hinges around the positivistic application of behavioural economics and its support of rationality (even if bounded). In a nuanced and culturally sensitive way, however, Clapham argues that there is a need to consider the behaviour of households in housing search:

“I accept fully Susan Smith’s plea for more empirical work. Rather than being based on simplified a priori assumptions of behaviour, frameworks need to be based on empirical findings of the behaviour of agents in specific markets.” (Clapham, 2011, P.291).

These frameworks need to account not only for an overview of search processes, but also variation in the behavioural processes between households.

Segmentation of supply and demand

Segmentation in housing analysis does not normally focus on demand; more routinely dissecting markets by supply, or dwelling characteristics (see Islam and Asami, 2009, for a fuller review of housing market segmentation). In these types of study hedonic regression analysis is frequently used to differentiate between submarkets based on property characteristics and price differentials (Goodman, 1978; Adair et al., 1996; Sheppard, 1999; Malpezzi, 2003), but these studies are based on assumptions of standardised demand. Some studies also segment households, but these are normally based on socio economic or demographic attributes rather than search processes or household economic behaviour (e.g. race: see Kain and Quigley, 1972, sexuality: see Ahmed et al, 2008; religion: see McPeake, 1998).

Brown and Moore (1970) argued that housing research needs to conduct surveys to explore the variation in the urban population’s housing needs and the variation in their search experiences (primarily the information sources used). Demand should be considered as segmented as households seek different attributes from the dwellings that they hope to purchase, and have different approaches to the search process (Megbolugbe et al., 1991; Gibler and Nelson, 1998). Demand segmentation may vary from area to area, as households’ cumulative preferences are non-uniform between locations (Munro and Lamont, 1985). Kaynak and Meidan (1980) found different preferences and consumer segments between Yorkshire, England (including Sheffield) and Nova Scotia, Canada, suggesting that there are cultural differences between attribute preferences. Within housing markets preferences vary across a wide range of household characteristics, including income, employment status and household composition (Tu, 2003). Households’ experiences of the housing market also differ and are likely to be reflected in the characteristics of their search (Anglin, 1997).

Studies into buyer search behaviour have also shown that a range of institutional market factors can influence the process, including market conditions (Baryla et al., 2000; Chernobai and Hossain, 2012), constraints in supply (Butler and Hamnet, 2012) and are shaped by real estate agent actions (Jud and Frew, 1986; Elder et al., 1999).

Gibler and Tyvimaa’s (2014) segmentation of consumer demand creates four categories based on life cycle stages, recreational activities and financial expenditures, suggesting that consumer demand is more varied than the new build dwellings currently developed. Feitelson (1993) segments owner-occupier demand using a hierarchical approach according to societal constraints, life-style choices and situation (stage in life cycle and income). These are two of the few examples of studies on consumer segmentation, but they are based on the premise that the search process plays little role in mediating inputs and outcomes of residential demand. One notable area of exception in consumer search segmentation is elderly migration and decision making, which has received a disproportionate amount of attention compared to the whole market. Wiseman and Roseman’s (1979) research sparked a cottage industry of gerontology housing search studies that reacted to the earlier premise that elderly households were considered a homogeneous group (e.g. Speare and Meyer, 1988), although many of these studies resorted to aspiration, preference or motivation based typologies, rather than behaviour or actions in the housing search process (e.g. Gibler and Taltavull, 2010). If elderly households are heterogeneous in their search behaviour, the whole population (all age groups) is likely to display greater variation than an elderly-only subset.

However, there has been relatively little research into the segmentation of behaviour in the housing search process. Piazzesi et al., (2014), in the only major consumer segmentation study based on search behaviour, found that in San Francisco households’ behaviours varied between more and less expensive neighbourhoods, with more expensive neighbourhoods being searched less often and properties there sold less frequently.

The lack of evidence of household variation in the housing search process has led to calls for more research.

“Analysis of household search processes can reveal key pressures and linkages within local markets and suggest where latent demands really exist. More work needs to be done in this area, not least in identifying the search patterns of different consumer groups and the extent to which consumers use hierarchical search processes: that is, establishing whether households focus on area, or type or some other attribute in selecting possible dwellings and then refocus on a second attribute and so on. Evidence already suggests that households have different and hierarchic search processes. For instance, whereas many Scottish households first select housing tenure, there is evidence that some younger households have strong area/ house-type preferences that dominate the tenure attribute. Some households may place house type and size ahead of area.” (Maclennan, 2012, P.20)

This thesis, therefore, seeks to make a contribution to the newly (re)energised applied behavioural research agenda in housing economics. It has an explicit demand side orientation and is intended, in particular, to explore the complexity of the search process of households who are purchasing a dwelling to live in1, in a way that draws on behavioural economics analysis. It also seeks to move beyond the assumption that behaviour is a uniform concept, explicitly drawing out the variation in different households’ search behaviours.

1.2 Aim of the study

The aim of this research is to create a typology of household search behaviour in the owner occupied sector in Sheffield.

The objectives of the study are:

  1. To review the literature on developing a conceptual model of housing search

  2. To create a conceptual model of the types of variation between housing search behaviours

  3. To test the variation in housing search behaviour through empirical research

  4. To devise a typology of housing search behaviour and relate this typology to the existing literature

  5. To critically assess the implications for applied research of the typology

Download 13.27 Mb.

Share with your friends:
1   2   3   4   5   6   7   8   9   ...   60

The database is protected by copyright ©sckool.org 2022
send message

    Main page