Who actually decides? Parental influence on the housing tenure choice of their children

Abstract

Chinese

We focus on the role of within-family socialisation and the relationship between socialisation and resource transfers in the intergenerational transmission of housing preferences, the formation of familial housing attitudes and thus the reproduction of a normative housing tenure ladder across generations in Czech society. We show that resource transfers and the within-family socialisation of housing preferences, including preferences concerning housing tenure, are closely interconnected. In other words, parental influence on decision to buy own housing (and on housing preferences in general) of their adult children through socialisation is stronger if there is an (actual or assumed) intergenerational resource transfer. This has several implications for how housing markets and systems work. The paper draws on findings from qualitative, quantitative and experimental studies.

Keywords

experiment housing homeownership resource transfers social networks socialisation

Introduction

Homeownership is the life goal of most households in advanced and transition countries. Although cultural differences do exist and are reflected in the different housing systems that have evolved (Kemeny, 1981), in many countries homeownership is more than just a ‘smart idea’ when compared pragmatically with the advantages and benefits of tenancy; it has become an imperative, a necessity and a norm.

There is extensive research, dating as far back as the 1980s, demonstrating that in many countries there exists a social norm that deems homeownership is superior to renting, despite the comparable security and costs of the two forms of tenure. Observations from various studies (Bazyl, 2009; Boehm, 1982; Coolen et al., 2002; Krumm, 1984) gave birth to the concept of a ‘housing career’ and, in stratification studies, to the concept of ‘housing class’, both of which consider housing tenures hierarchically on a normative ‘housing ladder’. Homeownership, and specifically ownership of a detached house, figures in many cultural environments at the top end of the ladder (see, for example, for the USA, Clark and Dieleman, 1996; Saunders, 1990; Shlay 2006; Taylor, 1999; for Australia, Bourassa, 1995; for Belgium, De Decker and Geurts, 2003; for the UK, Flint and Rowlands 2003; Lauster, 2010; for Denmark, Andersen, 2011; for Finland, Ruonavaara, 1996; for Japan, Hirayama, 2010).

In the countries of Central and Eastern Europe, socialist ideology thwarted the process whereby homeownership establishes itself as a social norm. Private ownership was officially suppressed. Nevertheless, it was common for detached homes to be privately owned (and in some countries, such as Bulgaria, the former Yugoslavia or Hungary, also apartments). Moreover, after 1990 the homeownership rate dramatically rose as a result of the mass give-away privatisation of the public housing stock. In many societies the homeownership rate exceeded 90% and such housing systems started to be referred to as ‘super-homeownership’ systems (Stephens, 2005).

In the Czech Republic, despite a more gradual increase in the homeownership rate, Lux and Sunega (2010) also demonstrated experimentally the strong public preference for homeownership. According to the Housing Attitudes 2013 survey, a national survey of the preferences and attitudes of the Czech adult population towards housing conducted in 2013, 83% of adult Czechs agree with the statement ‘living in your own home is always better than being a tenant, but not everyone can afford their own home’ (53% agreed strongly with this statement).¹ This paper seeks to describe the role of the family in reproducing the prevailing attitude (norm) that deems homeownership is always better than tenancy in Czech society. We will focus on socialisation of tenure (and other housing) preferences within family and will show that the formation of what we call familial housing attitudes is closely associated with within-family intergenerational resource transfers. To the best of our knowledge, this is the first time that the mutual link between socialisation relating to housing preferences and resource transfers has been empirically documented, i.e. that level of socialisation in housing tenure attitudes (norms) is closely associated with the existence or absence of within-family intergenerational resource transfers. This link has some significant implications for behaviour of agents on housing markets and housing system trends briefly outlined in the next section.

A review of existing research

An inconclusive discussion took place about whether a normative housing ladder, and the position of homeownership at its top end, is an expression of a natural human preference or need (Saunders, 1990), or whether this normative ranking is imposed by elites (Lauster, 2010) or some other visible (Gabriel and Rosenthal, 2005; Kemeny, 1978, 1981) or less visible power (Doling and Ford, 2003; Gurney, 1999). Saunders (1990) views the desire for homeownership as deriving from the natural need for ‘ontological security’ and as a solution to the problem of alienation in capitalist society. By contrast, Lauster (2010) sees the sources of the norm in the incentives of the privileged to distinguish themselves from the marginalised members of society. Gurney (1999) favours instead Foucault’s concept of the omnipotent and invisible ‘disciplinary’ power, in the exercise of which homeowners figure as both objects and subjects.

It is probably impossible to empirically rule out the plausibility of any of these theoretical concepts (see, for example, Mandic and Clapham, 1996): a human need for security may interact with power in a very complex manner that cannot be empirically explained. If one accepts that there are important methodological obstacles to determining the sources of normative behaviour, some progress may perhaps be made using a less ambitious approach, such as searching for the mechanism that reproduces the normative laddering of housing tenures in society. However, there are many channels of influence and they may again interact in a complex way, channels such as family, the media, reference groups and politics. Since adherence to a social norm occurs on an unconscious level, it may be difficult, if not impossible, to empirically distinguish the relative effect each channel has in maintaining a normative ranking of housing tenures in a society.

Parents certainly play a crucial role in the formation of preferences of their children;² therefore, and in an effort to ensure the feasibility of our research, we focused on a more thorough description of the mechanism of within-family housing preferences’ reproduction. The literature suggests two main channels through which parents can influence housing choices of their adult children (Bengtson and Roberts, 1991; Easterlin, 1987; Helderman and Mulder, 2007; Kohli, 1999; Mulder, 2007): resource transfers (Blaauboer, 2010; Mulder and Smits, 2013) and socialisation (Henretta, 1984). Empirical studies carried out to date have focused mainly on resource transfers, and socialisation effect remains empirically unexplored. The reason is that the studies used mainly quantitative survey data, which have no information on family members’ attitudes. However, it would be odd for intergenerational resource transfers to occur if the next generation did not share the attitude towards homeownership of the previous generation.

Easterlin (1987) hypothesises that the housing conditions a person becomes accustomed to in childhood may determine the baseline for housing preferences in later life. Based on this, Lersch and Luijkx (2015) operationalise socialisation as years spent in the parental home, but this kind of quantitative approach does not provide deeper empirical insight into the socialisation process itself. More thorough research of the (within-) family socialisation relating to formation of housing preferences can therefore bring new knowledge to research on housing tenure reproduction.

Socialisation is defined broadly as the process by which a person acquires attitudes, values, behavioural standards, world views, habits and skills (Quane and Wilson, 2012: 2978; Rowlands and Gurney, 2000: 122). This makes it difficult to grasp in analytically precise terms. Therefore, we need to start with a more precise framing of the term in our research. First, we will focus solely on the influence of parents on formation of children’s attitudes; and only on attitudes that concern housing. Second, we will understand socialisation as a process that occurs not just in childhood but also in later stages of the life cycle. Finally, socialisation may operate on the level of explicit verbal influence or through observation and the learning process. Consequently, we will focus both on (1) explicit (within-) family discussions, parental interference and advice, and (2) more implicit signs of long-term socialisation.

The main research question of this paper is: What is the relationship between (within-) family socialisation of housing preferences (especially those towards homeownership) and intergenerational resource transfers in Czech society? Our main proposition is that the (within-) family socialisation of housing preferences and resource transfer function in tandem, i.e. there is a close association between the existence of socialised (familial) housing attitudes and the existence (expectation) of intergenerational financial transfers. In other words, parental impact on the formation of housing preferences of their adult children and the intergenerational transmission of social norm relating to housing tenure through within-family socialisation is stronger if there is an (actual or assumed) intergenerational resource transfer and weaker if there is no transfer.

Our proposition implies that in families where the likelihood of a transfer is low the housing attitudes of the next generation are less familial and are formed more independently than vice versa; and, similarly, in families with a higher probability of transfers – mostly high-income families – the attitudes of young adults are significantly more familial, socialised and thus more ‘conservative’ than vice versa. This implication may have broader consequences for how the housing market works: there may be a longer-lasting normative inertia to home-buying behaviour (i.e. insensitivity to the changing market and other contextual conditions) in housing systems in which substantial family transfers are made than is observed in behaviour in systems in which family transfers are less significant. If (expected) resource transfers cause stronger socialisation, then any future interruption of transfers would significantly weaken the familial nature of housing attitudes and thus also weaken the within-family reproduction of the normative laddering of housing tenures. However, the broader market implications or cause–effect relations are not tested in this paper and remain open for future research.

The institutional context

We conducted our analysis in the Czech Republic where (1) there has been considerable growth in the share of owner-occupied housing during the last three decades as result of the privatisation of municipal housing; but where (2) the housing tenure structure is not so markedly skewed in the direction of owner-occupied housing and a substantial part of the housing stock is rented; and where (3) rents have been deregulated and therefore prices/rents are not distorted by state intervention. In 1991, the homeownership rate in the Czech Republic was 38%, co-op housing formed 19% and public rental 39% of the housing stock; private renting was almost non-existent. By 2011, the homeownership rate had increased to 56%, co-op housing made up 9%, public rental 8%, and private rental housing 14% of the housing stock.

The initial increase in private renting occurred in the early stages of the post-communist transition when apartment buildings that had been expropriated in the past by the communist regime were restituted to their owners or the owners’ heirs (see details in Lux and Mikeszová, 2012). In 1993 caps on rent were lifted from newly vacated flats (new tenancies), and soon after a large number of small landlords emerged in the market. The share of private rental housing out of the total housing stock consequently grew much faster than in advanced countries, going from almost zero in 1990 to 7% in 1993 (as a result of property restitution) and to 14% in 2011. Eventually a few institutional investors emerged in the Czech housing market; two of them became major private landlords in the local market.

The vast majority of rental tenancies recorded in the 2011 census were legal tenancies, i.e. based on a formal lease signed between landlord and tenant. The private rental market was stabilised by the rapid expansion of the housing stock and the deregulation of rents after 2007. While the average apartment price between 2000 and 2008 increased by almost 200%, the increase in average market rent in the same period was much more gradual – it grew by 64%. The value of the price-to-rent ratio increased; for example, in Prague, the national capital, it rose from 13.7 to 26.0. In 2008, in most regional capitals the average rent was below the level of the average user costs of homeownership (not including price appreciation). The market quickly changed from being supply-driven to demand-driven; and this fact led to improvements in the quality of housing and the security of private tenants in high and middle segments of the market. However, despite the increasing affordability and quality of private renting, popular attitudes significantly favour owner-occupied housing.

The acquisition of owner-occupied housing is largely accompanied by family aid: according to the survey Housing Attitudes 2013, 30% of homeowners acquired their housing through an inheritance or as a gift, and of those who bought their first home at market price or built it themselves, 42% received financial assistance (usually as a gift they do not have to pay back) from their parents. The biggest intergenerational assistance is observed among the youngest cohorts.

Methodology and data

In this paper we combine quantitative analyses of survey data with a qualitative study and an experiment. For the quantitative analyses, we used data from the Housing Attitudes 2013 survey on a sample of 3003 respondents.³ We tested for tenure status reproduction across generations and the role of intergenerational resource transfers in status reproduction using a binary logistic regression analysis.

The questionnaire survey did not include interviews among different family members (such as, parents and their adult children). Moreover, the socialisation of preferences affects more than just tenure preferences, but large quantitative surveys do not generally examine in detail different family members’ preferences regarding various housing attributes. Also, a shared preference for homeownership (or even similarity in other housing preferences) across generations within a family is not on its own sufficient to proxy for socialisation as it may simply reflect a similar reaction to the joint given market/policy situation. Finally, housing preference formation may be an unconscious process and, even if it does occur consciously, any direct questioning about family influence on preference formation is susceptible to survey biases such as priming, framing or cueing. Consequently, instead of a questionnaire survey we used an experiment, in which we searched for parental influence on the housing preference formation of adult children indirectly: we tested for an association between (a) the level of similarity in housing preferences within a family dyad composed of a parent and his/her adult child, and (b) the level of reciprocal knowledge measured on a scale indicating how well parents are able to guess the preferences of their children and how well adult children are able to estimate the housing preferences of their parents. Housing preferences were measured on 12 housing attributes and their combinations. The purpose was to test whether similarity in housing preferences across generations is associated with good reciprocal knowledge of these preferences across generations, because this association may better serve as a proxy for within-family socialisation than similarity of preferences alone.

We invited 82 subjects (mostly students who were expected to make their tenure choice in a few years) from a subject pool of participants in economic experiments. The participants were instructed to bring one of their parents with them to the experiment. The experiment involved letting 82 dyads of parents and their adult children evaluate sets of hypothetical housing profiles. The subjects were first asked to evaluate the profiles on their own and then to guess his/her partner’s housing preference. We designed two sets of hypothetical housing profiles. We refer to the first set of profiles as the ‘fundamental set’ and it included the following housing attributes: housing tenure (four categories), dwelling size (four categories), the dwelling’s construction material (two categories), technical state (three categories) and type of heating (three categories). The second set of housing profiles combined characteristics that relate to the living environment: namely, air quality, neighbourhood relationships, security, level of noise, access to green areas, access to the city centre and specific characteristic relating to workplace access; all of them two-category variables. This second set of profiles is the ‘environmental set’ of profiles.

We employed a fractional design⁴ so that the subjects would not have to evaluate all theoretically possible combinations of attribute levels. In fact, it was sufficient for the subjects to evaluate just 12 profiles of fundamental characteristics and 8 profiles of environmental characteristics (Table 1A in the Appendix). The subjects were asked to evaluate the hypothetical profiles on a scale of 1 to 20 points, where 1 is the best outcome and 20 the worst. We tried to define a methodological framework where the respondent would be incentivised to reveal their own housing preferences ‘truly’ and ‘honestly’. Both members of a dyad were given financial incentives to evaluate sets of hypothetical flat profiles as sincerely as possible so that their partner in a dyad could later replicate these ratings as accurately as possible. The financial payoff for each subject was based on each person’s ability to correctly estimate their partner’s evaluation of the hypothetical housing profiles. The financial incentives induced the respondents to engage in the mental effort of evaluating housing profiles; and providing precise estimates of the housing profiles of their partners.⁵

A second qualitative study sought to understand in-depth family interactions and the role the socialisation of housing attitudes plays in home-buying decisions. The study was conducted among 63 first-time buyers who recently decided to buy housing in one of two big Czech cities. As the study’s original purpose was to test the market implications of home-buying decisions, tenants were not included in the sample design. However, since the focus of our research is on the reproduction of the ethos of homeownership, these unique secondary data are also relevant. The informants took part in seven focus groups, which included role-playing in a simulated family conversation about tenure choice. In addition, in-depth interviews with eight dyads of respondent (first-time buyer) and the respondent’s parent were conducted. We applied a content analysis of the transcripts of the interviews and focus groups using Atlas.ti software. In the content analysis we combined inductive open coding and coding within selected categories, where the units of analysis were sentences and/or paragraphs.

The results from the experiment and from the qualitative study are not representative for the total population. Since most of our data are qualitative, we tested a qualitative proposition in this study, for which, however, limited generalisation is possible given the simultaneous use of quantitative data and an experimental study. The experiment, unlike the quantitative survey and the qualitative study, surveyed stated and not revealed housing preferences, but this does not detract from the value of its results because the purpose of the experiment was to examine socialisation in (whatever ideal) housing preferences within family. Table 1 presents the characteristics of the survey, the experiment and the qualitative study.

Table 1.

Characteristics of surveys and qualitative studies.

Survey/study	Housing Attitudes 2013	Housing attitudes across generations	First-time buyers panel
Method	Questionnaire survey, quota sampling	Experiment	Qualitative study (focus groups)
Size of sample	3003	164 (82 dyads)	63 (plus 8 parents)
Spatial coverage	Czech Republic, representative sample	Prague	Brno 62%; Ostrava 38%
Gender	Male 49%; female 51%	Male 37%; female 63%	Male 41%; female 59 %
Age	Average age 47 years	Average age of parents 51 years; average age of adult children 23 years	Average age 31 years
Education	University 16%; secondary 34%; less than secondary 50%	University 50%; secondary 27%; less than secondary 23%	University 81%; secondary 18%; less than secondary 2%
Marital status	Married 50%; cohabiting 12%; single 17%; divorced 13%	N/A	Married 22%; cohabiting 46%, single 24%, divorced 7%

Findings

We used data from the Housing Attitudes 2013 survey to test for tenure status reproduction across generations within the family and the role of intergenerational resource transfers in this status reproduction. In the survey, the respondents – parents with direct tenure title and at least one child aged 18+ years – were asked to give the age, marital status, place of residence and housing tenure of each of their adult children. This question was answered by 919 respondents who provided data on 1178 adult children.

Our main hypothesis is that socialisation of housing preferences within the family is closely associated with transfer-giving. Although we could not measure the socialisation of preferences in this questionnaire survey, our hypothesis implies that tenure status reproduction should be associated with transfer-giving. In other words, adult children will be more likely to reproduce their parents’ homeownership status if they receive or can expect to receive a transfer of resources from their parents, and will be less likely to do so if they have no such assistance. We tested whether (1) housing tenure is reproduced across generations (after controlling for other relevant factors) and (2) whether this reproduction weakens when transfer-giving is controlled for in the analysis. We created a logit model where the housing tenure of the adult child was a dependent dummy variable with the following values: (1) owner and (2) other tenure. The independent variables were gender, age and the marital status of the adult child; household income and housing tenure of the respondent-parent; and a dummy variable indicating whether the respondent-parent had ever given or was planning to give financial support to buy a home to at least one of his/her adult children. Model 1 in Table 2 shows that children whose parents are homeowners have a much greater chance of also being homeowners, after controlling for other factors. When the variable on transfer-giving is included (Model 2), the reproduction of housing tenure status weakens as expected by our hypothesis. However, this weaker reproduction of tenure status may be a sign not just of weaker intergenerational transmission of housing preferences through socialisation but also of housing unaffordability: homeownership may not be affordable for adult children who do not receive any parental resource transfer.

Table 2.

Reproduction of housing tenure between parents (respondents) and their adult children – logit model coefficients [Exp(B)].

Factor / model		Model 1		Model 2
Variable	Category	Exp (B)	Sig.	Exp (B)	Sig.
Gender (child)	Female	0.912	0.498	0.932	0.630
	Male (ref.)
Age (child)		1.066^***	0.000	1.060^***	0.000
Marital status (child)	Married	4.215^***	0.000	4.024^***	0.000
	Not married (ref.)
Tenure (parent)	Owner	2.157^***	0.000	1.692^***	0.006
	Cooperative member	1.171	0.539	0.935	0.808
	Other (ref.)
Household income (parent)	Higher income	1.581	0.164	1.190	0.630
	Middle income	1.417^**	0.015	1.250	0.151
	Lower income (ref.)
Financial transfer (parent to child)	Yes	.	.	8.931^***	0.000
	No (ref.)
Intercept		0.032 ^***	0.000	0.030 ^***	0.000
Model descriptives:
	N	1178		1178
	−2 Log likelihood	1310.7		1142.3
	Cox & Snell R square	0.228		0.331
	Nagelkerke R square	0.306		0.444
	Corr. predict (%)	72.7		76.0

Notes: The asterisks represent significance at a 1% significance level (***), a 5% significance level (**), and a 10% significance level (*). The dependent variable is the housing tenure of the adult child: (i) owner-occupancy; (ii) other tenures. Ref. indicates the reference category in the estimated model. Income was self-categorised by the respondent. The model coefficients are the odds ratio [Exp(B)], where values >1 indicate a higher probability of the particular choice and, conversely, values <1 indicate a lower probability. Independent variables were checked for collinearity.

Source: Housing Attitudes 2013.

A qualitative study of socialisation

As another step towards answering our research question we conducted an experiment. Recognising the methodological limits of research that concerns stated preferences, we tested for any links between (a) the similarity of housing preferences within a family dyad, composed of a parent and his/her adult child, and (b) family members’ reciprocal knowledge of each other’s preferences, measured on a scale indicating how well parents are able to guess the housing preferences of their children and how well adult children are able to guess the preferences of their parents. The association between the level of reciprocal knowledge and the similarity of preferences serves as a more robust sign of within-family socialisation than the mere similarity of preferences across generations.

For each family, we estimated a ‘family’ model using the data from the profile evaluations made by parents and their offspring. We took advantage of the fact that everybody evaluated the same sets of profiles on the same rating scale. We used the individual-level ratings of the housing profiles within each of the categories – fundamental and environmental. We constructed a model in which we pooled the profile evaluations for each dyad of a parent and his/her offspring. We thus constructed 82 ‘family’ models using the following specification:

P r o f i l e_r a t i n g_{i p} = α + \sum_{a = 1}^{A - C} A t t r i b u t e_l e v e l_{a i p} * β_{a} + ε_{i p} .

(1)

The dependent variable in model (1) is a rating of housing profile p by individual i, where i={1,2}, i.e. being either a parent or his/her offspring. The evaluated profiles fall either within the fundamental or the environmental set of profiles. The independent variables are the set of (A–C) variables for the particular housing-attribute levels present in profile p, where A is the total number of attribute levels and C is the total number of attribute categories. The coefficients represent the average part-worth utility contributions of the respective attributes within the family. We applied OLS to estimate model (1).

Following Molin et al. (2001), we interpret the model fit (R²) of each estimated model (1) as a measure of the closeness of preference functions between the parent and his/her offspring. The higher the model fit (R²), the closer the preferences of the respective family members. If there are identical housing preferences across the generations, the model fit will be perfect (R² = 1). On the other hand, we based the payoff earned in our experiment on the parents’ ability to guess the preferences of their children and the children’s ability to guess the preferences of their parents. Hence, we argue that the larger joint payoff earned in our experiment provides a proxy for a deeper reciprocal knowledge of housing preferences. Analogically, the larger parent payoff is a proxy for a parent’s better knowledge of the housing preferences of his/her adult child and the larger offspring payoff is a proxy for an offspring’s better knowledge of the housing preferences of his/her parent.

We examined the correlations between the model fit (similarity in preferences) and the different types of experimental payoffs (reciprocity in knowledge of preferences). Table 3 shows that there is a significant correlation (at the 10% significance level) between reciprocal knowledge (joint payoff) and a similarity of preferences within a family for both sets of profiles. However, when reciprocal knowledge is measured solely according to how well parents know the preferences of their children (parent payoff, columns 3 and 4) the association between knowledge and similarity of preferences was weak. On the opposite, when reciprocal knowledge was measured according to how well children know the preferences of their parents (children payoff, columns 5 and 6) the association between knowledge and similarity of preferences was even stronger than in the case of joint reciprocal knowledge.

Table 3.

Association between reciprocal knowledge of housing preferences and the similarity of housing preferences within the family: Correlation coefficients and OLS estimates between the joint preference model fit (R²) and the experimental payoff (joint, parent and offspring).

	(1)	(2)	(3)	(4)	(5)	(6)
	Model fit in fundamental set of profiles	Model fit in environmental set of profiles	Model fit in fundamental set of profiles	Model fit in environmental set of profiles	Model fit in fundamental set of profiles	Model fit in environmental set of profiles
Joint payoff	0.11*	0.12	–	–	–	–
	(0.06)	(0.08)
Parent payoff	–	–	0.12	0.14	–	–
			(0.10)	(0.10)
Offspring payoff	–	–	–	–	0.28**	0.25
					(0.12)	(0.16)
Corr. coeff.	0.20* (p = 0.07)	0.19* (p = 0.08)	0.15 (p = 0.19)	0.16 (p = 0.16)	0.23** (p = 0.04)	0.19* (p = 0.09)
N	82	82	82	82	82	82

Notes: Six models are presented in this table. The dependent variable is either the joint preference model fit for the fundamental set of profiles or for the environmental set of profiles. Each model has only one regressor: the joint payoff for the family (in thousands of CZK), the parent payoff (in thousands of CZK) or the offspring payoff (in thousands of CZK), respectively. The asterisks represent significance at a 1% significance level (***), a 5% significance level (**), and a 10% significance level (*). OLS estimates are presented with robust standard errors in parentheses. Correlation coefficients are reported with the appropriate significance levels.

Source: Housing Attitudes across Generations, experiment.

Consequently, we did not find any proof that similarity of housing preferences is the result of explicit (within-) family discussion and communication. It is true that the more informed adult children are about the housing preferences of their parents the more their preferences will be similar to those of their parents. However, as parents were much less informed about their adult children’s housing preferences than vice versa, and good parental knowledge of their children’s preferences was not significantly associated with a similarity of housing preferences across generations in this case, the similarity of housing preferences within a family seems to be more the result of implicit socialisation and the socialised factors of the lived environment than of explicit discussions within family.

An experiment gave us another important finding. We selected families (a) with an above-median joint preference model fit (R²), used to measure the similarity of the preferences between the parent and his/her offspring and, simultaneously, (b) that received above-median joint (or offspring, respectively) payoffs, which is used here as a proxy indicator for reciprocal knowledge; and then we compared them with the rest of our sample. The results (middle columns of Table 4) show that (socialised) families observed to have strong implicitly familial housing attitudes likely also have significantly higher earnings.⁶ It is obvious that parents from upper-income families are more likely to live in the housing they prefer; consequently, their children are also better at predicting their preferences. However, the preferences of adult children were also more similar to the preferences of their parents in this case, which is probably a sign of the existence of familial (socialised) housing attitudes in this social group. If we look at the similarity of preferences between parents and their offspring only (last column of Table 4), we find no significant socioeconomic difference.

Table 4.

Differences in income and education between families with strong and weak familial attitudes.

		Familial attitudes (joint payoff)		Implicitly familial attitudes (offspring payoff)		Similarity of preferences
		Strong familial attitudes^a	Weak familial attitudes^b	Strong familial attitudes^c	Weak familial attitudes^d	High similarity of preferences^e	Low similarity of preferences^f
Fundamental characteristics	Parental	30,211.54*	25,084.54 (13,235.89)	30,040.00**	24,811.12 (13,035.67)	27,834.98	25,585.37 (13,121.02)
	income (in CZK)	(p = 0.07) (16,076.18)	25,084.54 (13,235.89)	(p = 0.03) (16,295.40)	24,811.12 (13,035.67)	(p = 0.24) (15,470.14)	25,585.37 (13,121.02)
	University-educated parents (in %)	50 (p = 0.55) (0.51)	52 (0.50)	60 (p = 0.14) (0.50)	47 (0.50)	56 (p = 0.19) (0.50)	46 (0.50)
	Observations	26	56	25	57	41	41
Environmental characteristics	Parental income (in CZK)	30,925.00* (p = 0.07) (15,162.52)	25,350.55 (13,864.19)	30,521.74** (p = 0.07) (17,026.37)	25,224.31 (12,946.22)	27,798.39 (p = 0.25) (14,806.85)	25,621.95 (13,871.18)
	University-educated parents (in %)	50 (p = 0.55) (0.51)	52 (0.50)	57 (p = 0.28) (0.51)	49 (0.50)	46 (p = 0.81) (0.50)	56 (0.50)
	N	20	62	23	59	41	41

Notes: The numbers with standard errors in parentheses represent the means of the presented variables. The p-values in parentheses report the one-sided mean comparison t-test comparing the differences in the tested variables between the groups within each box. The asterisks represent significant differences at a 1% significance level (***), a 5% significance level (**), and a 10% significance level (*).

Dyads characterised by above joint preference median model fit (R²) and above-median joint payoff.

Dyads characterised by below joint preference median model fit (R²) and below median joint payoff.

Dyads characterised by above joint preference median model fit (R²) and above-median offspring payoff.

Dyads characterised by below joint preference median model fit (R²) and below median offspring payoff.

Dyads characterised by above joint preference median model fit (R²).

Dyads characterised by below joint preference median model fit (R²).

Source: Housing Attitudes across Generations, experiment.

The experiment gave us three findings: (1) there seems to be little explicit (within-) family discussion of housing preferences between parents and children, but (2) familial socialised attitudes do indeed exist (in the sense that implicit socialisation or living-environment socialisation leads children to follow their parents’ attitudes), and (3) they are more likely to be found in families with higher income where the resource transfers are also larger.⁷ In the subsequent qualitative study of first-time buyers we made a deeper examination of (within-) family communication and modes of socialisation relating to tenure preferences and home-buying. We analysed the transcripts of role-playing that was carried out in seven focus groups with 63 first-time buyers, the transcripts of in-depth interviews and data from mini-questionnaire survey⁸ conducted among 38 first-time buyers and transcripts of eight in-depth interviews with the parents of first-time buyers.

In each focus group, respondents were asked whether their parents had influenced their decision to buy housing or not. Respondents who answered in the affirmative formed one group and they played the role of parents; the other respondents formed a second group and they played the role of an adult child. The respondents in the first group were asked to replicate the arguments of their parents during within-family communication and to try to persuade the respondents in the second group to buy housing; the respondents in the second group were supposed to oppose them. We analysed the use of arguments by the respondents playing the role of parents: at what point in the discussion the argument was raised, how frequently the arguments were used and their (persuasive) impact (Table 5). We assume that the most frequent arguments and the arguments that appeared at the beginning or the end of the role-playing activity are the ones that the respondents regarded as important. In addition, the impact of the argument was valued based on whether the counterparty (the respondents who role-played the children) was able to come up with a counter-argument. If they were unable to find a meaningful response, the impact was assumed to be strong. The table shows that the promise of financial aid (resource transfer) is regarded both as the most important argument and the one with the strongest impact (the discussion often ends after it is raised).

Table 5.

The position and impact of arguments of ‘parents’ in the role-playing activity in the focus groups.

		Where the argument occurred in the discussion			Impact of the argument
Argument	Occurrences	Beginning	Middle	End	Discussion continues	Discussion ends (win)	Discussion ends (lose)
Financial assistance to buy a home is offered	10	6	1	3	5	4	1
Ownership represents security and a good investment	9	4	5	0	9	0	0
Some other form of assistance towards buying a home is offered	8	3	2	2	5	1	2
Homeownership is currently financially advantageous	6	2	4	0	5	1	0
Homeownership is always financially advantageous	6	1	3	2	4	1	1
A person should own a home at a certain age	5	1	2	2	2	2	1
A person should own a home once he/she has a stable job	3	3	0	0	3	0	0
A person should own a home when they are planning a family	3	2	1	0	3	0	0
Renting is insecure and uncertain	3	0	2	1	2	1	0

Source: First-Time Buyers Panel. N=68 informants, 7 focus groups.

In the second stage, we drew on data from mini-questionnaires, where first-time buyers were asked to name whom they had consulted with when they were deciding whether or not to buy a home and estimate how important consultation with each person was in their decision.⁹ We compared the results for two groups of respondents: those who either have received or have been promised a resource transfer from their parents and those who have not. As Figure 1 shows, parents are much more important consultants when home-buying is accompanied by an actual (or planned) resource transfer than vice versa. The difference between these two groups of respondents is in fact quite large.

Figure 1.

Influence of reference groups in home-buying decision of first-time buyers. Question: ‘We are interested in learning who you consulted with on home-buying and who is of the greatest help to you in the home-buying decision’. The numbers indicate the ‘average’ importance of particular reference groups in home-buying decisions separately for each of the two groups of respondents: with and without resource transfers. The importance of the reference group is measured as the average distance between the individual frames and the centre of the diagram (representing the respondent) using AutoCAD design software.

In the third stage, we analysed 16 interviews of eight dyads (parents and their adult children) that were selected non-randomly based on the voluntarily decision of respondents to invite their parents to take part in the research. The interviews were conducted separately. In the interviews we found that parents often stress their ‘non-interference’ in their children’s decisions and, similarly, their adult children tended to stress the ‘autonomy’ of their decision-making. However, in the same breath, parents added that they provided their children with advice, consultation or financial support. This seeming contradiction was observed in all the parental interviews. For all children in the dyads, the parents played a significant role: they provided advice and financial help (gifts, savings) and offered to serve as the lenders of last resort. In all cases, parents and children shared the same tenure preference, and in only three cases did they disagree on particular issues (timing, locality, type of housing). Familial housing attitudes thus coexist with ‘non-interference’ and ‘autonomy’ in the home-buying decision itself. The statements in the interviews confirmed that subtle, gradual and long-term implicit socialisation proved to be the most effective means by which shared familial housing attitudes are formed. An illustration of this is provided by the following excerpt from an interview with one parent:

Well, we’ve never given them advice, because you shouldn’t counsel children too much. You should maybe just guide them a little bit. But he has always been a responsible person. […] Well, it’s hard to describe how to guide them, I think you do it by raising them in a certain way, not telling him: ‘You must buy a house’. (Pauline, 8:23)

The parents, all of them homeowners, simply regard homeownership as the best possible way of resolving one’s housing situation and they pass this ethos on to their children through the socialisation process as a whole (which includes the effect of socialisation through the lived environment). The decisions their children make about tenure then retain the formal appearance of free and independent decisions. During socialisation, parents tend to use symbols and values, such as the notion that homeownership provides security in its broadest sense,¹⁰ offers real independence (freedom to furnish one’s home as one wishes), establishes a real home for a family (providing children with roots and a sense of belonging), and it is a sign of responsibility and adulthood (achievement in life). Similarly, studies on social reproduction in general confirm that parents want their children to be able to at least maintain the same social status that they have (see, for example, Ortner, 1998).

The strategy of performative ‘non-interference’ but de facto interference resembles a popular metaphor: two people are standing on their own boats on a lake, and if they throw each other a ball and their interaction is face-to-face and direct, the boats will move further apart, but if they stand back to back, throw a boomerang instead of a ball, and their interaction is more indirect, their boats will move closer together. The subtle and delicate forms of influence that parents exercise on their children’s decisions through (a) non-explicit long-term socialisation into certain sets of moral and cultural values and through (b) indirect respectful conversation have a much more effective influence on the decisions their children make than would forcing them to make a choice they do not want.

Finally, we analysed the relationship between such long-term socialisation of housing preferences and resource transfers in the sample of 38 first-time buyers. We identified long-term socialisation when (1) a first-time buyer explicitly mentioned long-term family influence/discussion concerning housing; or (2) made reference to family housing history as influential for the formation of his/her own housing preferences; or (3) stated that it is ‘natural’ for children to have housing preferences similar to those of their parents. We found that resource transfers were received by 20 out of 25 respondents among whom there was an evidence of long-term socialisation, while only by 5 out of 13 respondents whose interviews revealed no evidence of long-term socialisation.

Conclusions

In this paper we focused on the role of the socialisation of housing preferences and the link between this socialisation and resource transfers in the intergenerational transmission of housing preferences, the formation of familial attitudes towards homeownership, and thus the reproduction of the normative housing tenure ladder in Czech society. Our proposition was that intergenerational resource transfers and (within-) family socialisation of housing preferences are interconnected. In other words, parental influence on the formation of their adult children’s housing preferences and the intergenerational transmission of housing tenure social norms are stronger if there is a (real or planned) intergenerational resource transfer and vice versa.

Given the nature of our data (research on socialisation is complex and methodologically challenging), we were unable to directly and representatively prove our qualitative proposition, but we found several indirect pieces of evidence supporting it. First, the within-family reproduction of housing tenure status across generations weakens when resource transfers are controlled for. Second, the strongest signs of the (within-) family socialisation of housing preferences and the existence of implicit familial housing attitudes were, in our experiment, observed among higher income families, in which there are also bigger resource transfers. Finally, we showed that (1) the promise of financial aid likely has the strongest impact on the home-buying decisions of young first-time buyers during intra-family communication and (2) the role of parents as consultants was found to be much more significant for the decision to buy a home if their children also received (would receive) a resource transfer than if they did not. We also showed that long-term implicit socialisation, which is more effective at transmitting housing preferences within the family, is more commonly found among first-time buyers who expect to receive or have already received a resource transfer from their parents than among those who have not received and will not receive any such transfer. However, the results from experiment and qualitative study cannot be generalised to the whole population.

A possible explanation for the link between transfers and socialisation is that children who receive material transfers may feel more confident and secure. While taking on a large mortgage debt may add to a person’s doubts about the rationality of purchasing a home, material transfers effectively reduce these doubts. If (expected) resource transfers cause higher socialisation of a norm, then were such transfers to cease or decrease in the future (for example, owing to the ageing of society and the increasing use of private welfare in retirement) the familial nature of housing preferences and the within-family reproduction of normative tenure laddering would be weakened. This could pave the way for a more balanced housing system where tenancy is not as stigmatised as it is now. Conversely, if parents, in particular in high-income families, continue to provide their adult children with resources the way they do it now we will instead see the consolidation of the normative housing ladder in Czech society. In the latter case we can expect growing housing tenure inequalities and a stronger link between social and housing tenure inequalities in a society where so far this link has been relatively weak (see, for example, Lux et al., 2013).

The giveaway privatisation of municipal housing that occurred and continues to occur in the Czech Republic (though in a much more decentralised manner and more gradually than in most other post-socialist societies) meant that households that could not have afforded to become homeowners under market conditions have acquired homeownership status. As the housing stock that remains to be privatised decreases and eventually disappears, the combined effect of a strong norm and an unequal distribution of family resources available for transfers may lead to increasing housing inequalities in the society. However, the broader housing market and system implications and the question of cause and effect remain open for future research.

Footnotes

Appendix

Table 1A.

Fundamental set of profiles.

	Housing tenure	Flat type and floor space	Heating	Technical state of the house/flat	Construction material
1	Ownership with a 4% IR mortgage	2 rooms + kitchen / 70 m²	Electric heating	Older building / renovated flat	Brick
2	Ownership with a 4% IR mortgage	2 rooms + kitchen / 70 m²	Gas heating	Older building / renovated flat	Brick
3	Ownership with a 4% IR mortgage	3 rooms + kitchen / 60 m²	Off-site heating	Older building / renovated flat	Concrete
4	Ownership with an 8% IR mortgage	3 rooms + kitchen / 60 m²	Electric heating	New building / new flat	Brick
5	One-year rental contract	2 rooms + kitchen / 50 m²	Gas heating	Older building / renovated flat	Concrete
6	One-year rental contract	3 rooms + kitchen / 60 m²	Off-site heating	New building / new flat	Brick
7	One-year rental contract	3 rooms + kitchen / 80 m²	Electric heating	New building / new flat	Concrete
8	Ownership with an 8% IR mortgage	3 rooms + kitchen / 80 m²	Gas heating	Older building / unrenovated flat	Concrete
9	Rental contract for an indefinite period	3 rooms + kitchen / 80 m²	Off-site heating	Older building / unrenovated flat	Brick
10	Rental contract for an indefinite period	2 rooms + kitchen / 70 m²	Electric heating	Older building / unrenovated flat	Concrete
11	Rental contract for an indefinite period	2 rooms + kitchen / 50 m²	Gas heating	Older building / unrenovated flat	Brick
12	Ownership with an 8% IR mortgage	2 rooms + kitchen / 50 m²	Off-site heating	New building / new flat	Concrete

Source: Housing Attitudes Across Generations, experiment.

Acknowledgements

We are grateful to Tomáš Dvořák for valuable research assistance.

Funding

The research for this paper was sponsored by the Czech Science Foundation (Grantová Agentura České Republiky) with grant number 16-06335S.

Notes

References

Andersen

(2011) Motive for tenure choice during the life cycle: The importance of non-economic factors and other housing preference. Housing Theory and Society 28: 183–207.

Bazyl

(2009) Factors influencing tenure choice in European countries. Central European Journal of Economic Modelling and Econometrics 1: 371–387.

Bengtson

Roberts

REL

(1991) Intergenerational solidarity in aging families: An example of formal theory construction. Journal of Marriage and Family 53(4): 856–870.

Blaauboer

(2010) Family background, individual resources and the homeownership of couples and singles. Housing Studies 25(4): 441–461.

Boehm

(1982) A hierarchical model of housing choice. Urban Studies 19: 17–31.

Bourassa

(1995) A model of housing tenure choice in Australia. Journal of Urban Economics 37(2): 161–175.

Clark

Dieleman

(1996) Households and Housing: Choice and Outcomes in the Housing Market. New Jersey: Rutgers.

Coolen

Boelhouwer

Van driel

(2002) Values and goals as determinants of intended tenure choice. Journal of Housing and the Built Environment 17: 215–236.

De Decker

Geurts

(2003) Belgium. In: Doling

Ford

(eds) Globalisation and Home Ownership. Delft: Delft University Press, pp. 21–52.

10.

Doling

Ford

(eds) (2003) Globalisation and Home Ownership. Delft: Delft University Press.

11.

Easterlin

(1987) Birth and Fortune: The Impact of Numbers on Personal Welfare. Chicago, IL: University of Chicago Press.

12.

Flint

Rowlands

(2003) Commodification, normalisation and intervention: Cultural, social and symbolic capital in housing consumption and governance. Journal of Housing and the Built Environment 18(3): 213–232.

13.

Gabriel

Rosenthal

(2005) Homeownership in the 1980s and 1990s: Aggregate trends and racial gaps. Journal of Urban Economics 57: 101–127.

14.

Gurney

(1999) Lowering the drawbridge: A case study of analogy and methaphor in the social construction of home-ownership. Urban Studies 36(10): 1705–1722.

15.

Helderman

Mulder

(2007) Intergenerational transmission of homeownership: The roles of gifts and continuities in housing market characteristics. Urban Studies 44(2): 231–247.

16.

Henretta

(1984) Parental status and child’s home ownership. American Sociological Review 49(1): 131–140.

17.

Hensher

(2006) How do respondents process stated choice experiments? Journal of Applied Econometrics 21(6): 861–878.

18.

Hirayama

(2010) The role of home ownership in Japan’s aged society. Journal of Housing and the Built Environment 25: 175–191.

19.

Kemeny

(1978) Forms of tenure and social structure: A comparison of owning and renting in Australia and Sweden. The British Journal of Sociology 29: 41–56.

20.

Kemeny

(1981) The Myth of Home Ownership: Private Versus Public Choices in Housing Tenure. New York, London: Routledge.

21.

Kohli

(1999) Private and public transfers between generations: Linking the family and the state. European Societies 1(1): 81–104.

22.

Krumm

(1984) Household tenure choice and migration. Journal of Urban Economics 16: 259–271.

23.

Lauster

(2010) Housing and the proper performance of American motherhood, 1940–2005. Housing Studies 25(4): 543–557.

24.

Lersch

Luijkx

(2015) Intergenerational transmission of homeownership in Europe: Revisiting the socialisation hypothesis. Social Science Research 49: 327–342.

25.

Lux

Mikeszová

(2012) Property restitution and private rental housing in transition: The case of the Czech Republic. Housing Studies 27: 77–96.

26.

Lux

Sunega

(2010) The future of housing systems after the transition: The case of the Czech Republic. Communist and Post-Communist Studies 43(2): 221–231.

27.

Lux

Sunega

Katrňák

(2013) Classes and castles: Impact of social stratification on housing inequality in post-socialist states. European Sociological Review 29(2): 274–288.

28.

Mandic

Clapham

(1996) The meaning of home ownership in the transition from socialism: The example of Slovenia. Urban Studies 33(1): 83–97.

29.

Molin

EJE

Oppewal

Timmermans

HJP

(2001) Analyzing heterogeneity in conjoint estimates of residential preferences. Journal of Housing and the Built Environment 16(3–4): 267–284.

30.

Mulder

(2007) The family context and residential choice: A challenge for new research. Population, Space and Place 13(4): 265–278.

31.

Mulder

Smits

(2013) Inter-generational ties, financial transfers and home-ownership support. Journal of Housing and the Built Environment 28(1): 95–112.

32.

Ortner

(1998) Identities: The hidden life of class. Journal of Anthropological Research 54(1): 1–17.

33.

Quane

Wilson

(2012) Making the connection between the socialisation and the social isolation of the inner-city poor. Urban Studies 49(14): 2977–2987.

34.

Rowlands

Gurney

(2000) Young peoples? Perceptions of housing tenure: A case study in the socialization of tenure prejudice. Housing, Theory and Society 17(3): 121–130.

35.

Ruonavaara

(1996) The home ideology and housing discourse in Finland 1900–1950. Housing Studies 11(1): 89–105.

36.

Saunders

(1990) A Nation of Home Owners. London: Unwin Hyman.

37.

Shlay

(2006) Low-income homeownership: American dream or delusion? Urban Studies 43(3): 511–531.

38.

Stephens

(2005) A critical analysis of housing finance reform in a ‘super’ home-ownership state: The case of Armenia. Urban Studies 42(10): 1795–1815.

39.

Taylor

(1999) Places, spaces and Macy’s: Place–space tensions in the political geography of modernities. Progress in Human Geography 23(1): 7–26.

40.

Timmermans

Molin

Noortwijk

(1994) Housing choice processes: Stated versus revealed modelling approaches. Journal of Housing and the Built Environment 9(3): 215–227.