An Institutional Economic Analysis of the Decision to Do-it-yourself in Housing Renovation

Abstract

This paper examines home renovators’ decision in choosing do-it-yourself (DIY) or hiring contractors. DIY may reduce financial pressure but may lead to potential domestic discord and perils of incompetence; hiring contractors may alleviate time pressures but may require more financial resource. This choice is modelled as analogous to Williamson’s corporate strategic decision between ‘outsourcing’ and ‘internalisation’ in the context of vertical integration, and extended via Richardson’s capabilities-based theory, with social reputation effect taken into account. Based on a survey of Brisbane home renovators, it is found that the Richardson’s thesis seems stronger than the Williamson’s when applied to housing renovation and that ‘trust’ issues in respect of contractors are paramount.

1. Introduction

In Australia, renovators’ decisions to do-it-yourself (DIY) potentially reduce their financial pressure and help them to enjoy a better property despite declining housing affordability, which in 2007 dropped to the lowest since 1984 (Kathy 2007). However, the attempt to save money by opting for DIY renovation may have adverse impacts on renovators’ well-being due to heightened time pressure, foregone leisure (see Kahneman and Krueger, 2006) or disrupted daily lives (Hooton, 2006; Masia and Miller, 2006; Wyld, 2005). Another downside of DIY renovation are safety hazards due to inexperience; for example, Hoffman (2006) reports that in the Australian state of Victoria there are 15 renovation-related deaths each year and more than 500 hospital admissions with 33 per cent of eye injuries being related to DIY activities.

Such factors would be the natural points of focus if following the economics literature on renovation, which presents a utility-maximising model to explain renovators’ choice between ‘hiring others’ or ‘DIY’. Here, however, an attempt is made to examine such decisions from the perspective of institutional economics with input from behavioural economics and socioeconomics.

The paper is structured as follows. The theoretical framework is presented in the next section, followed by a section discussing data collection. The fourth section provides descriptive statistics of surveyed renovators. The model, hypotheses and empirical results of renovators’ likelihood of supplying their own labour are presented in the fifth, sixth and seventh sections respectively, to see if, besides conventional factors, renovators’ DIY decision is further driven by factors similar to those that drive corporate internalisation decisions (Williamson, 1985; Richardson, 1972) and by social/psychological attitudes. The final section presents the conclusion.

2. Theoretical Framework

In the economics literature, decision-making about the use of one’s labour in renovation is portrayed as a process in which well-informed individuals maximise their utility and productivity subject to budget and time constraints (Bogdon, 1996; Mendelsohn, 1977; Pollakowski, 1988). This paper, by contrast, follows Earl’s (1986, 2005) argument and views the choice between ‘hiring others’ and ‘DIY’ as analogous to the corporate strategic decision between ‘outsourcing’ and ‘internalisation’. Vertical integration theory has been applied to ‘make or buy’ decisions involving complex sub-contracting in the construction industry (Chau and Walker, 1994; Earl, 1996; Walker, 2007; Walker and Chau, 1999), but potential for applying it to domestic property renovations has hitherto largely been ignored. Moreover, the present paper adds further perspectives from behavioural economics and socioeconomics to examine the extent to which psychological factors interfere with renovators’ strategic decisions.

Decisions in this context involve several characteristics that have not been taken into account in the previous studies. Firstly, renovation often entails a production chain of many procedures, requiring various types of expertise. A renovator does not simply face a simple one-shot choice between DIY and hiring a contractor, but a set of choices about who will undertake different types of tasks during the project.

Secondly, renovation is a creative process, full of uncertain choices challenging renovators’ rationality. Cognitive constraints may cause renovators to struggle when trying to work out the right combination of inputs in housing production. For example, given that people have trouble keeping more than seven things in mind at a time (Miller, 1956), they may be puzzled by the task of choosing the best combinations from an array of 2000 available colours of paint. Worse still, renovators’ plans may be derailed, not merely by possibilities they had imagined but not taken seriously, but also by unexpected surprises such as the discovery of termite problems once the project is underway (see Shackle 1972).

Thirdly, if bounded-rational renovators are indeed overwhelmed by too many choices, they may use social interaction to simplify the task (Earl and Potts 2004). While this may result in some renovators outsourcing work to contractors, it may interest first-time renovators in embarking on DIY projects on the basis of ‘rosy’ but false impressions—for example, via watching TV ‘makeover’ shows. Moreover, such decision-making could end up being driven by a social network effect (Bikhchandani et al., 1992; Earl et al., 2007); for instance, observing your neighbours’ DIY activities may make you want to give DIY a try (Phipps, 1983).

Fourthly, DIY renovation may be an outlet for excitement to escape the boredom caused by high standards of safety and security in the modern life (Scitovsky, 1981) and a source of fulfilment (Brooks and Earl, 1987). Hence, a renovator may choose DIY for psychological reasons instead of as a result of careful cost–benefit calculation that would see the challenging problems entailed in DIY as a cost, not a benefit.

Fifthly, renovation activities tend to be either experience goods whose quality can only be ascertained ex post, or worse still, credence goods where ex post assessments of quality are problematic when undertaken by hired labourers (Ekelund et al., 1995; Nelson, 1970). This leaves the scope for suppliers to engage in opportunism (Williamson, 1975, 1985), the self-serving behaviour that exploits information asymmetry. Given this, renovation can be viewed as a co-ordination game where trust is a far more important factor than conventional drivers of choice such as income, prices or productivity (Granovetter, 1985, 2005). A renovator’s decision to undertake DIY may be caused by his lack of trust in contractors although hiring others is optimal given his own skill limitation.

Finally, renovation is an irreversible conjunctive process, the value of which will be low if it is abandoned before completion. Renovators may feel compelled to finish the work at all costs because of project irreversibility, thin markets for torn out materials and the inconvenience an incomplete project causes in daily life (Peng, 2009). Besides raising the possibility of overcapitalisation (Peng, 2011), the conjunctive nature of renovation also induces contractors to engage in guileful behaviour.

Since renovation projects involve a sequence of processes akin to those in a firm’s supply chain, Earl (2005) argues that renovators’ choice between getting works done in the market (i.e. hiring others) and in-house (i.e. DIY) may be usefully analysed via Williamson’s (1975, 1985) transaction cost theory of corporate vertical integration. This theory portrays internalisation as being chosen by firms as a means of reducing potential for market failure and avoiding the transaction cost of getting things done in conditions where transactions are conducive to suspicion and fear.

Williamson (1985) predicts that buyers will become their own suppliers (i.e. internalise production) due to high transaction costs if all the following four conditions prevail

If decision-makers suffer from bounded rationality in a complex transaction environment.

If the other party to a transaction enjoys information advantage and is expected to engage in opportunism via pursuing their self-interest in a guileful manner by, for example, exploiting ambiguous parts of a contract.

If a situation of small numbers prevails on the other side of the transaction, ‘hold-up’ threats could be a problem since it will be difficult to dispose of such threats by making alternative arrangements.

If a production process involves asset specificity, the potential costs of an unsuccessful transaction are increased by the low possibilities of recouping prior investments by selling equipment or employing specially acquired skills on other projects.

Bounded rationality matters if there is opportunism, but opportunism would be prevented by the presence of large numbers of alternative trading partners, while even in a small-numbers case opportunism would not be a worry in the absence of asset specificity (i.e. if a deal falls apart, the invested asset could be resold or used elsewhere). Hence, in the presence of these four conditions, contracts cannot guarantee efficient transaction and thus production internalisation is likely to be the preferred strategy.

However, Earl and Wakeley (2005) argue that the issues of capabilities and trust may inevitably make market contracting a preferable choice even in the face of Williamson’s four conditions for a high risk of market failure. Richardson’s (1972) emphasis on capabilities leads to the view that, if Williamson’s four conditions prevail, it may still be better to outsource the work because internalisation without relevant skills may produce worse outcomes and/or higher costs. Besides, Granovetter (1985) claims that transaction is not always based on evaluation of transaction costs but rather, in some situations, bonded by trust, which discourages potential malfeasance and makes reciprocal negotiation a likely solution when disputes arise. A contractor’s misconduct and opportunistic behaviours are deterred by the possibilities of losing businesses in the long run by gaining an infamous reputation transmitted through his client’s social network.

From the perspective of behavioural economics, one would further recognise the scope for renovators’ psychological attitudes to derail rational decision-making when choosing between hiring others and DIY. Even if Williamson’s four conditions do not imply a case for DIY, a renovator without appropriate capabilities may still choose to use his own labour because psychological elements may overshadow rational judgement. For example, those DIY-renovators who got injured could have been driven by ‘renovators’ delight’. Similarly, an unjustified lack of confidence could lead ‘capable’ others to hire contractors to perform the tasks, which generates higher opportunity costs than using self-supplied labour. It is not proposed here that psychological factors will out-power conventional factors. Instead, in some situations, psychological factors as intervening variables may be powerful enough to outweigh conventional influence (Katona, 1960).

3. Data Collection

A questionnaire was sent to 3000 households across suburbs in Brisbane between November 2006 and January 2007. This questionnaire consisted of 43 questions about households’ renovation experience, renovation expenditure, renovation methods, demographics, and social and psychological attitudes. The data for 14 psychological attitudes were then quantified by factor analysis into four psychological variables—’perceived self-performance in renovation’ (perception of one’s own performance in undertaking renovation), ‘trust in contractors’ (trust in contractors one works with), ‘social dependence’ (concern about what others think) and ‘perceived ease of renovation’ (perception of how easy renovation is).

In all, 269 renovators (those with renovation experiences, out of 366 respondents) were dichotomised into 207 hire-renovators and 62 DIY-renovators in terms of renovation methods. ‘Hire-renovators’ are those who delegated the whole renovation project to one principal contractor (41 per cent of a total of 269 renovators), several contractors (15 per cent) or those who assigned most of the works to contractors and did the rest themselves (21 per cent); ‘DIY-renovators’ are those who did most of the work by self-supplied labour and left the rest to the hired labour (10 per cent), or those who did all the work by themselves unless a tradesperson was legally required for the task (13 per cent).

4. Differences Between DIY-renovators and Hire-renovators

4.1 Reasons for renovation and for Renovation Method

For both (surveyed) hire-renovators and DIY-renovators, the main reason for renovation was functional need (47 per cent of hire-renovators; 45 per cent of DIY-renovators), followed by lifestyle pursuit (23 per cent; 24 per cent). In the multiple-choice section of reasons for renovation strategies, hire-renovators chose to hire others because hired labours’ work guaranteed better quality (70 per cent), shorter working time (60 per cent), creative design (38 per cent) and cheaper quotes for inputs (26 per cent). DIY-renovators chose DIY due to having relevant skills (85 per cent), seemingly easy work (24 per cent), affordable tools (35 per cent) and difficulty in finding contractors in time (6 per cent) and within budget (16 per cent).

4.2 Demographics and Social/Psychological Factors

More than 75 per cent of hire-renovators and DIY-renovators were born in Australia. Compared with DIY-renovators, hire-renovators were older ( ${Mean}_{Hire}$ = 51.4; ${Mean}_{DIY}$ = 48.5) and more likely to have tertiary education (58 per cent of hire-renovators; 48 per cent of DIY-renovators) and higher annual household income (33 per cent of hire-renovator households with more than A$130 000; 25 per cent of DIY-renovator households). Although renovators had, on average, two wage-earners in the family, ( $M_{Hire}$ = 1.65; $M_{DIY}$ = 1.84), up to 16 per cent of DIY-renovators (7 per cent of hire-renovators) had three to four wage-earners in their household, implying that they were more financially independent and physically DIY-capable adults. DIY-renovator households tended to spend more time in getting to work per day ( $M_{DIY}$ = 50.3 minutes; $M_{Hire}$ = 47.1). Hire-renovators had older houses (house age: $M_{Hire}$ = 49.7; $M_{DIY}$ = 46.4) and had lived longer in their homes (length of residence: $M_{Hire}$ = 10.9; $M_{DIY}$ = 9.1).

In their recent renovation projects, hire-renovators had higher estimated renovation costs ( $M_{Hire}$ = A$58 938; M_DIY = A$18 556) and actual renovation costs ( $M_{Hire}$ = A$71 863; $M_{DIY}$ = A$21 254); hire-renovators spent more time on preparation for renovation (more than 4 months for 43 per cent of them and 23 per cent for DIY-renovators), implying that the size of hire-renovators’ projects tended to be bigger and more complex; hence, they are more likely to have asked for more quotes before renovation ( $M_{Hire}$ = 2.3; $M_{DIY}$ = 1.5).

Regarding renovation capability, DIY-renovators had more renovation experiences ( $M_{DIY}$ = 2.58; $M_{Hire}$ = 1.97). As for the type of renovation experience, hire-renovators had more experience in hiring others, while DIY-renovators had more experience in DIY.

Regarding psychological attitudes towards renovation, although both groups showed a similar level of perceived self-performance in undertaking renovation ( $M_{Hire}$ = 0.70; $M_{DIY}$ = 0.68), hire-renovators had higher trust in contractors ( $M_{Hire}$ = 0.59; $M_{DIY}$ = 0.25), and cared less about others’ opinions (social dependence: $M_{Hire}$ = −0.74; $M_{DIY}$ = −0.60). Yet undertaking a renovation project seemed to be harder to hire-renovators (perceived ease of renovation: $M_{Hire}$ = −0.33; $M_{DIY}$ = −0.06). Both groups were similar in their tendency to take adventurous holidays ( $M_{Hire}$ = 0.43; $M_{DIY}$ = 0.40), which is used here as a proxy for their demand for excitement.

5. Likelihood of DIY Job: Model and Data Screening

This paper’s key aim is to examine renovators’ likelihood of engaging in DIY. The dependent variable is binary (DIY-renovator, D = 1; hire-renovator, D = 0) with the probability (p) as the expected value. Logistic regression is used to model p indirectly via logit transformation of p as $\ln [p / (1 - p)]$ . $[p / (1 - p)]$ is known as the odds ratio, indicating the ratio of the probability of an event occurring—being a DIY-renovator in this case—to the probability of it not occurring (Tabachnick and Fidell, 2007).

\begin{matrix} \ln [p / (1 - p)] = α + β_{1} X_{1} \\ + β_{2} X_{2} + β_{3} X_{3} + \dots + β_{k} X_{k} \end{matrix}

with $\ln [p / (1 - p)]$ , known as log-odds, as the dependent variable; and X are independent variables.

The log-odds of renovators’ likelihood of choosing ‘DIY’ is modelled as a function of 16 independent variables, whose hypotheses will be elaborated in the next section. Before empirical analyses, the data screening of all independent variables satisfied the assumptions of normality, homogeneity of variance, linearity and multicollinearity, but the reliability of logistic regression is sensitive to outliers. Since several multivariate outliers (cases with standardised residuals $\geq | 2.58 |$ at p = 0.01; see Garson, 2008) were detected, two regressions—one with intact data and the other without outliers—will be presented in section 7.3 to see if the influence of any independent variable is disguised by the existence of outliers.

After removing the cases with missing values, the total sample size (255: 197 hire-renovators and 58 DIY-renovators) is sufficient for the requirement of logistic regressions with 16 and 17 independent variables (Norman and Streiner, 2000) presented in the following sections.

6. Likelihood of DIY Job: Hypotheses

In order to probe renovators’ likelihood of using their own labour (DIY), several hypotheses (represented by the relevant variables derived from the survey) are elaborated as follows.

6.1 Fundamental Factors

Seven variables are the proxies for opportunity costs and marginal productivity. ‘Income’ as a proxy for opportunity cost of self-supplied labour, and ‘education’ as a proxy for permanent wage rate are assumed to be negatively associated with renovators’ inclination towards DIY (Bogdon, 1996; Potepan, 1989).

A renovation project is a time-consuming process, in which a renovator may need to ‘live in, eat at and sleep on renovation’ for months (Hooton, 2006; Jamrozik, 1994). Thus, the variable ‘time spent (by all wage-earners) in travelling to work each day’ is to measure time availability because time shortage during weekdays will also impose time pressure during weekends (for renovation).

Aspects of household composition that affect marginal productivity in renovation include ‘number of wage-earners’ (i.e. the number of physically capable adults in DIY) and ‘age of renovator’ (Bogdon, 1996). While the effect of ‘number of wage-earners’ on DIY likelihood is assumed to be positive, the effect of ‘age of renovator’ is uncertain in the economics literature. Nevertheless, this study assumes a positive effect of ‘age of renovator’ on DIY because older people in Australia nowadays: enjoy good health supported by an advanced medical system; have more time available after retirement; and, tend to choose DIY for general fitness enhancement and satisfaction in a job well done (Ashby et al., 2007).

Moreover, structural factors approximate the marginal productivity of a house itself: ‘age of house’, implying the degree of wear and tear, and ‘length of residence’, indicating the likelihood of wear and tear that need to be addressed (even if the house was in pristine condition at the time tenants moved in), are assumed to be negatively related to DIY likelihood, for an old and possibly long-neglected property may require a huge investment in labour input to bring it up to modern standards.

6.2 The Risk of Contractual Failure

Based on Williamson’s (1975, 1985) and Richardson’s (1972) theories of internalisation in corporations, the hypotheses regarding the choice between DIY (internalisation) and hiring contractors (outsourcing) are as follows.

Bounded rationality and complexity

A renovation project as a production chain with high uncertainty is intensified by the tendency of renovation services being experience goods or credence goods. This element assumes that renovators should internalise the work if the complexity of a situation makes it hard to negotiate deals with contractors that offer a guaranteed end-result at a specified price. This may be due to fears of complications caused by ambiguous and unforeseen circumstances such as the content of a task getting ‘lost in translation’ when what a renovator has in mind is not the same as what the contractor pictures as what his client wants (even if the contract is carefully worded and the contractor tries to comply with it). This problem of uncertainty tends to be in positive proportion to the size of a project; in line with Williamson’s theory, as this variable rises, DIY will be preferred.

Opportunism

Claims of having specialist expertise can be used by tradespeople as a cover for self-serving behaviour at their clients’ expense. A good example is found in Herbert Simon’s reports of his encounter at Illinois Tech with architecture faculty members, who believed that

the client was to be educated, persuaded … to contribute the resources necessary to produce a great work of art … The client was an instrument, a means (Simon, 1991, pp. 164–165).

With a bigger project involving things likely to be out of control, hired labour is vulnerable to the ‘sin of information’—“experts can keep information asymmetry as asymmetrical as possible” (Levitt and Dubner, 2005, p. 69) and thus is less likely to “let his conscience bother him” (see Vandiver, 1972, for a vivid example of plane construction). Williamson’s perspective thus implies that the risk of opportunism rises with the size of a project.

Few alternative supplies (small numbers)

From a mainstream viewpoint, a great number of potential contractors listed in the yellow pages indicates that the problem of ‘small numbers’ seems unlikely. However, finding contractors can sometimes be more difficult than finding ‘a needle in a haystack’ if they are in short supply but in strong demand—as in Australia in recent years (Colvin, 2005; Kruger, 2007; Mares, 2008). DIY seems more attractive when renovators are aware of supply shortage for they cannot readily count on contractors doing a good job for fear of being fired or suffering reputational damage due to bad ‘word of mouth’ reports (May, 2005). With a short supply of contractors, further incentives to favour DIY arise if a renovator’s job is relatively small, resulting in contractors being unwilling to supply a quotation.

The ‘number of quotations’ that renovators got prior to renovation is used as a proxy to capture their initial transaction costs in finding contractors willing to provide a quotation and do the job within the required time-frame.

Specificity of equipment and immobility of renovation asset

In Williamson’s analysis of corporate vertical integration, asset specificity may make firm-to-firm transactions problematic. Suppliers dislike investing in specific assets that are to serve a particular customer because difficulties in reselling or reusing such assets put them in a vulnerable position afterwards; downstream customers will be wary of investing in specific assets if upstream suppliers are likely to engage in a ‘hold-up’ strategy to renegotiate the contract once the investment has been made.

In the case of renovation projects, tool specificity is not really an intractable problem for contractors even if they install the items before receiving payment, since the risk of customers’ refusing to pay is normally reduced via the use of instalment payments. Instead, asset specificity is best seen as arising due to the conjunctive dimension of renovation: if a contractor breaks a contract and leaves the project messily unfinished, the customer will face major costs in getting it back on track and materials already used and paid for may be stuck in situ. Such non-recoverable investments put renovators in a vulnerable position if they use contractors to do the work, for the latter may, in a bid to extract higher fees, threaten to walk off the job and thereby impose major dislocation costs on renovators. In the face of this possibility, DIY is a possible way to ensure that the non-recoverable investments will be brought swiftly to fruition. The bigger the project, the more this asset specificity/conjunctive dimension militates against taking risks with contractors, so the probability of opting for DIY on this count will be a function of the size of a renovation project.

Summary: Williamson’s four conditions

While summarising Williamson’s four conditions, three of them—bounded rationality/uncertainty, opportunism and specificity—are related to the size of a renovation project in terms of the range of involved tasks as the measure of its complexity and risk. Three variables were considered to approximate this factor: ‘estimated renovation costs’, ‘types of works involved in a renovation project’ and ‘time spent on preparation prior to renovation’. The first one seems plausible but actually unreliable since a large outlay might reflect a lot of the same kind of activity (as with roofing a property as opposed to undertaking plumbing, painting, tiling and electrical work in a bathroom renovation) or indicate the use of materials with higher quality and prices.

The second one is an ideal proxy, derived from a survey question regarding the types of work involved in a renovation project. However, renovators’ unclear or even blank replies to this question can only generate a broadly defined dummy variable ‘structural alteration/addition or not’ (project involved structural changes of alteration and addition = 1; otherwise = 0) as a proxy for the size of a renovation project. Due to the roughness of this dummy variable, another variable ‘time spent on preparation prior to renovation’ is considered, which indicates that the more time spent on preparation (getting quotes from contractors, arranging time between workers, buying materials, etc.), the larger the size of the renovation and the bigger the concern about being let down by contractors or having disputes with them and hence, from the standpoint of Williamson’s theory, ultimately the less likely a renovator is to hire others to do the work (i.e. internalisation through DIY will be preferred). A significant positive Pearson correlation (0.326, p <.001) between ‘structural alteration/addition’ and ‘time spent on preparation’ may imply ‘time spent on preparation’ is a reasonable proxy for the size of a renovation project: a renovator tends to spend more time on preparation prior to renovation if a project involves a structure-changing alteration/addition.

Hence, three of the Williamson’s factors (bounded rationality/uncertainty, opportunism and specificity) will be approximated by the size of renovation, which is captured by ‘time spent on preparation’ in empirical regressions, while a supplementary regression using ‘structural alteration/addition’ instead will also be presented for comparison. Moreover, the issue of small numbers will be captured by the ‘number of quotes’ that renovators asked for prior to renovation.

Richardson’s argument: capability

The renovation process requires various different forms of expertise (installing a tap, hanging a door, etc.). Following Richardson’s (1972) argument, if a renovator is not equipped with relevant skills, outsourcing his renovation project is the only solution even at the risk of uncertain outcomes, opportunistic malfeasance and lack of substitute contractors. Richardson’s capabilities perspective thus leads to the opposite prediction from Williamson’s transaction cost perspective in this context. The indivisibility of tools also favours outsourcing in large and complex projects because the limitations of the tool rental market may compel DIY-renovators to buy expensive tools that they do not expect to use again. However, if a renovator has accumulated a diverse range of capabilities through experience, he may choose DIY even if Williamson’s four conditions are not present.

‘The number of renovation experiences’ seems to be a good proxy for a renovator’s capabilities but what really matters may be the type of renovation experience. It is assumed that, even when not being threatened by Williamson’s four conditions, a renovator is more likely to engage in DIY if he has prior DIY experiences that help him to deal with possible problems during DIY renovation, whereas a renovator with experiences in hiring others will be more likely to continue to do so, having the capability to get good deals and having competence in supervising the projects. Two generated dummy variables—’experience in hiring others’ and ‘experience in DIY’—represent the type of renovation experience. In order to see if renovators’ choice between methods is more likely to be driven by the type of renovation experience or by the number of renovation experiences, two empirical results will be presented for comparison.

6.3 Social and Psychological Factors

The set of social and psychological variables is included to explore whether a renovator’s decision to engage in DIY is being shaped by factors that subjectively anchor his perceptions of the situation differently from how a mainstream economist would see it. Under the social/psychological influences, some renovators who could financially afford to outsource the work may still opt for DIY even in the absence of Williamson’s four conditions for internalisation.

Trust and social dependence

Granovetter (1985, 2005) argues that trust is a key element that forges interaction and transaction between individuals. If a renovator trusts potential contractors, he may decide to outsource even if Williamson’s conditions for contractual failure seem to hold. Although trust is generally developed through on-going relationships, renovators (especially first-time renovators) may at some stage choose to hire others simply because of their good impression of contractors, which is built upon the image received from people in their social network (such as neighbours), or even by using simple decision rules such as ‘You can judge a tradesman by the state of his truck’ following a tradesperson’s visit to prepare a quotation (see Alpers and Gerdes, 2006). Thus, the variable ‘level of trust in contractors’ is assumed to be negatively related to the likelihood of DIY (while positively with the likelihood of hiring others).

Given the possible role of social interaction in the process of generating trust, the variable ‘level of social dependence’ is used to see if a renovator who cares what others think may want to give people a good impression by using his own labour in renovation.

Overestimation of one’s capability

It has been shown in psychological experiments that individuals fail to recognise their incompetence especially in the areas where they lack experience (Dunning et al., 2003). The variable ‘perceived self-performance in renovation’, which presents a renovator’s evaluation of his own capabilities regardless of his actual experiences in DIY, is thus assumed to be positively related to DIY likelihood.

Underestimation of DIY task

Renovators may choose DIY simply because they underestimate uncertainty and possible physical danger entailed in the DIY process (Bedi and Goldbloom, 2008; Phipps, 1983; Routley and Ozanne-Smith, 1995). Although it is widely reported that most DIY-renovators without skills suffer from consequences such as accidents/injuries, ruining materials, damaging an adjacent part of a house (Masia and Miller, 2006), their initial impression may be partly anchored by numerous popular Australian TV renovation shows in which a dilapidated house can be revitalised in 30 minutes (Dunlevy, 2003; Finnerty, 2007). Therefore, DIY likelihood is assumed to be positively related to a renovator’s ‘level of perceived ease of renovation’.

Pursuit of adventurous challenges and impact of nationality

Besides looking for consistent and novelty-limited comfort (i.e. safety and security in life), individuals also desire extraordinary excitement out of their daily routine. In Scitovsky’s (1981, 1992) analysis of well-being, the first boredom-free source he lists is DIY activities, followed by gambling and dangerous sports such as river rafting (studied by Arnould and Price, 1993) or skydiving (studied by Celsi et al., 1993). The variable ‘tendency to take adventurous holidays’ as a proxy for renovators’ tendencies towards excitement is assumed to be positively related to DIY likelihood.

Australia nowadays has a highly multicultural population. Although DIY renovation is particularly regarded as a way of pursuing excitement by local Australians (Jamrozik, 1994), immigrants with different cultural backgrounds may think otherwise. Whereas Americans, British and Australians regard DIY as a hobby, the main difficulty that foreign hardware shops (such as B&Q from Britain) have in enlarging their businesses in China is their preference for BIY (buy-it-yourself) since “middle-class Chinese feel it is beneath them to build a cabinet or fit a shelf” (The Economist, 2003). The dummy variable ‘born in Australia’ (D = 1, yes) is generated to approximate how people culturally view using their own labour.

These hypotheses and relevant variables are summarised in Table 1.

Table 1.

Hypotheses and relevant variables: likelihood of DIY renovation

Hypotheses	Variables
Fundamental factors
1. The higher a renovator’s family income, the less likely he chooses DIY	‘Level of household income’
2. The higher a renovator’s education level, the less likely he chooses DIY	‘Level of education’
3. The older a renovator, the more time he has and thus the more likely he chooses DIY	‘Age of renovator’
4. The more the number of wage-earners in the household, the more likely a renovator chooses DIY	‘Number of wage-earners’
5. The older a house, the less likely a renovator chooses DIY	‘Age of house’
6. The longer the residence, the less likely a renovator chooses DIY	‘Length of residence’
7. The more time spent in travelling to work, the less likely a renovator chooses DIY	‘Time spent in travelling to work’
Contractual and capability risk factors
8. Williamsons’ theory: the larger the size of a renovation project, the higher the possibilities of uncertainty, opportunistic behaviours and asset specificity, the more likely a renovator chooses DIY (versus Richardson’s theory: the larger the size of a project, the more likely a renovator will outsource the project to overcome capability deficits)	‘Time spent on preparation’ prior to renovation (in models 2-1, 2-2)
	Note: alternative variable ‘structural alteration/addition’ is also considered (in model 2-3)
9. The harder it is to find a contractor, the fewer quotes a renovator can get prior to renovation, so the more likely a renovator chooses DIY	‘Number of quotes’
10-1. The more renovation experiences a renovator has, the more likely he chooses DIY	‘Number of renovation experiences’ (in models 2-1, 2-3)
10-2 (a) If a renovator has experience in DIY, he is more likely to DIY	‘Experience in DIY’ (in model 2-2)
10-2 (b) If a renovator has experience in hiring others, he is more likely to hire others (i.e. less likely to DIY)	‘Experience in hiring others’ (in model 2-2)
Social and psychological factors
11. The higher a renovator’s trust in contractors, the less likely he chooses DIY	‘Trust in contractors’
12. The more a renovator cares about others’ opinions, the more likely he chooses DIY	‘Social dependence’
13. The higher a renovator’s perceived self-performance in renovation, the more likely he chooses DIY	‘Perceived self-performance in renovation’
14. The higher a renovator’s perceived ease of renovation, the more likely he chooses DIY	‘Perceived ease of renovation’
15. The higher a renovator’s tendency to take adventurous holidays, the more likely he chooses DIY	‘Tendency to take adventurous holidays’
16. If a renovator is born in Australia, he is more likely to choose DIY	‘Born in Australia’

7. Empirical Results

In order to see if, on top of fundamental variables, contractual and social/psychological variables could further explain the variation in the DIY likelihood, these two blocks of variables were entered in sequence. In this sequential setting, two types of regression (models) are presented in section 7.1: the first one with the block of fundamental variables, the second one with all the variables (the block of fundamental variables and the block of contractual and social/psychological variables). Comparison of the pseudo R² results contributed by each regression indicates the extra variation in DIY likelihood that is explained by the added sets of contractual and social/psychological variables.

Moreover, extra attention is paid to the ‘experience’ factor (a proxy for Richardson’s ‘capability’ argument). Two sequential logistic regressions are thus presented in the following section: the first one (model 2-1) with the aggregate number of renovation experiences (i.e. the number of renovation projects a renovator had prior to his latest renovation) and the second one (model 2-2) with the type of renovation experience (i.e. the type of experience—hiring others or DIY—a renovator had prior to his latest renovation). Please note that the details of model 2-3 will be discussed in section 7.2.

7.1 Likelihood of DIY Renovation

The empirical results are listed in Table 2. As seen in model 1 with only the block of fundamental variables, this block of variables significantly accounted for 12 per cent of variation in DIY likelihood.

Table 2.

Dependent variable:0 = hire-renovator	Model 1			Model 2-1			Model 2-2			Model 2-3
Dependent variable:0 = hire-renovator	Fundamental variables only			All the variables			All the variables			All the variables
1 = DIY-renovator				Use (a) ‘time spent on preparation’			Use (a) ‘time spent on preparation’			Use (b) ‘structural alteration/addition’
				Use (c) ‘number of renovation experiences’			Use (d), (e) ‘type of renovation experience’			Use (c) ‘number of renovation experiences’
Independent variables	B	Exp(B)	Wald	B	Exp(B)	Wald	B	Exp(B)	Wald	B	Exp(B)	Wald
Block 1: fundamental variables
Age of house	−0.01	0.99	1.05	0.00	1.00	0.00	0.00	1.00	0.38	−0.00	1.00	0.03
Length of residence	−0.04	0.96	2.58	−0.03	0.97	1.27	−0.05	0.95	3.26 *	−0.03	0.97	1.03
Age of renovator	−0.03	0.97	2.88*	−0.04	0.96	4.84**	−0.04	0.96	3.49*	−0.04	0.96	4.74**
Level of education	−0.23	0.79	2.07	−0.28	0.75	2.20	−0.46	0.63	4.37**	−0.27	0.76	2.11
Level of household income	−0.16	0.85	3.37*	−0.13	0.88	1.44	−0.04	0.96	0.12	−0.12	0.88	1.47
Number of wage-earners	0.51	1.67	4.97**	0.42	1.52	2.87*	0.64	1.89	5.43**	0.38	1.46	2.23
Time spent in travelling to work	0.00	1.00	0.34	0.00	1.00	0.12	0.00	1.00	0.21	−0.00	1.00	0.10
Block 2: corporate and social/psychological variables
Size of a renovation project
(a) Time spent on preparation	—	—	—	−0.54	0.59	6.26***	−0.60	0.55	6.58***	—	—	—
(b) Structural alteration/addition	—	—	—	—	—	—	—	—	—	−0.01	0.99	0.00
Capability
(c) Number of renovation experiences	—	—	—	0.07	1.08	0.56	—	—	—	0.12	1.13	1.70
(d) Experience in hiring others (yes = 1)	—	—	—	—	—	—	−1.47	0.23	12.21***	—	—	—
(e) Experience in DIY (yes = 1)	—	—	—	—	—	—	2.52	12.46	21.65***	—	—	—
Number of quotes	—	—	—	−0.27	0.77	4.40**	−0.19	0.82	1.84	−0.33	0.72	6.70***
Level of trust in contractors	—	—	—	−1.04	0.35	12.00***	−1.13	0.32	10.46***	−1.07	0.34	13.05***
Level of social dependence	—	—	—	0.51	1.67	2.59	0.43	1.54	1.50	0.39	1.48	1.67
Level of perceived self-performance in renovation	—	—	—	0.14	1.15	0.07	0.07	1.08	0.02	0.03	1.03	0.00
Level of perceived ease of renovation	—	—	—	0.85	2.33	6.59***	0.42	1.52	1.25	0.83	2.30	6.59***
Tendency to take adventurous holidays				−0.14	0.87	0.08	−0.29	0.75	0.26	−0.18	0.83	0.13
Place of birth (Australia = 1)	—	—	—	0.04	1.04	0.01	0.35	1.42	0.41	0.08	1.09	0.03
(Intercept)	1.51	4.52	1.74	4.33	76.24	7.97	3.73	41.49	4.51	3.29	26.87	5.01
Model evaluation	Model 1			Model 2-1			Model 2-2			Model 2-3
Chi-squared in omnibus test	20.45***			63.67***			101.75***			56.64***
Pseudo R-squared	0.12			0.34			0.51			0.31
The effect of added block	Fundamental variables			Other variables			Other variables			Other variables
Chi-squared in omnibus test	20.45***			43.22***			81.26***			36.83***
Classification rate (percentage)	Model 1			Model 2-1			Model 2-2			Model 2-3
Hire-renovators correctly classified	77.2			81.7			83.7			82.2
DIY-renovators correctly classified	44.8			62.1			77.6			57.9
Overall	69.8			77.3			82.3			76.8

Notes: * p < 0.10; ** p < 0.05; *** p < 0.01.

Model 2-1 presents the results that the block of contractual and social/psychological variables—including the number of renovation experiences to capture the ‘capability’ factor—further added on top of fundamental variables. Model 2-1 significantly accounted for 34 per cent of variation in the likelihood of DIY renovation. The significant omnibus test of the added block of contractual and social/psychological variables ( $λ_{9}^{2}$ = 43.22) implies that this set of factors explained more variation in the DIY likelihood than the block of fundamental variables alone.

By contrast, model 2-2 presents the results that the block of contractual and social/psychological variables—including instead two dummy variables to capture the type of renovation experience (‘hiring others’ and ‘DIY’)—added on top of the block of fundamental variables. As seen in Table 2 regarding the overall evaluation, model 2-2 with the type of renovation experience performs better in terms of significantly accounting for 51 per cent of variation in the DIY likelihood, compared with 34 per cent in model 2-1 with the number of renovation experience. The added block of contractual and social/psychological variables was statistically significant ( $λ_{10}^{2}$ = 81.26) in explaining the DIY likelihood in model 2-2.

The evaluation of independent variables is also shown in Table 2. Only the final models (model 2-1 and model 2-2) with all the variables are discussed here. In model 2-1, several variables were statistically significant in explaining the DIY likelihood. Fundamental variables include ‘age of renovator’ (p <0.05) and ‘number of wage-earners’ (p <0.10), corporate variables include ‘time spent on preparation’ (p <0.01) and ‘number of quotes’ (p <0.05), and social/psychological variables include ‘trust in contractors’ (p <0.01) and ‘perceived ease of renovation’ (p <0.01).

The availability of self-supplied labour (‘number of wage-earners’) makes renovators more likely to opt for DIY. The opposite impact (to what is expected) of ‘age of renovator’ may imply that physical fitness remains a key factor although older homeowners have more free time to enjoy the satisfaction derived from DIY. Moreover, the influence of the size of renovation (‘time spent on preparation’) on the DIY likelihood contradicts the assumption based on Williamson’s transaction cost theory. This, on the other hand, is well explained by Richardson’s theory that the bigger the size of a project (i.e. involving a wider range of different capabilities), the more likely a renovator will outsource the project to overcome his capability deficits (i.e. the smaller the size of a project, the more likely a renovator will opt for DIY). This is also related to the fact that DIY-renovators tended to get fewer quotes before renovation. It should be noted that the number of renovation experiences used as a proxy for capability was not statistically significant.

Furthermore, compared with hire-renovators, DIY-renovators tended to have lower trust in contractors and to regard renovation as an easy and enjoyable process, which may partly explain why the number of DIY injuries rose in the past decade in Australia.

When it comes to model 2-2 where the capability factor is measured in terms of the type of renovation experience, Richardson’s theory still seems to hold: the significant negative effect of the size of a renovation project (‘time spent on preparation’) on the DIY likelihood may be attributed to renovators’ capability deficits. The significant influences of type of renovation experience may support this relationship (‘experience in DIY’, p <0.01; ‘experience in hiring others’, p <0.01): a renovator is likely to use his labour (hire others’ labour) in renovation when having experiences in DIY (in hiring others). Hence, the type of renovation experience matters more than the number of renovation experiences in measuring ‘capability’ in the context of renovation.

Fundamental variables are still influential in model 2-2. The likelihood of DIY renovation was positively related to ‘number of wage-earners’ (p <0.05), negatively related to ‘level of education’ (p <0.05), ‘age of renovator’ (p <0.10) and ‘length of residence’ (p <0.10). While ‘trust in contractors’ still played a crucial role in explaining the DIY likelihood (p <0.01), ‘perceived ease of renovation’, which was significant in model 2-1, was not significant in model 2-2—possibly because the variation that this variable accounted for overlaps with the variation that is explained by other variables.

To sum up, four variables significantly explain the DIY likelihood in both model 2-1 and model 2-2: ‘age of renovator’, ‘time spent on preparation’, ‘number of wage-earners’ and ‘trust in contractors’. Moreover, the type of renovation experience is a better proxy than the number of renovation experiences in measuring renovators’ capability. Renovators are more likely to opt for DIY (holding other factors constant)

— if they are younger;

— if they have more wage-earners in their household;

— if they spend less time in preparation (implicitly assuming the smaller size of a project);

— if they have a lower level of trust in contractors; and

— if they have prior experience of DIY activities.

7.2 A Note on the Size of a Renovation Project

As discussed earlier, ‘structural alteration/addition’ was an ideal proxy for the size of a project but was not used due to respondents’ unclear or even blank answers to this question. ‘Time spent on preparation’ was chosen instead for its significant positive correlation with ‘structural alteration/addition’, which implies that a project involving structural changes (and thus various tasks requiring different capabilities) will require more time in preparation. Another regression with ‘structural alteration/addition’ (model 2-3) is thus further presented for comparison in Table 2.

While comparing models 2-1 and 2-3 in Table 2, both variables that represent the size of a project (‘time spent on preparation’, ‘structural alteration/addition’) demonstrate a negative effect on the DIY likelihood, which supports Richardson’s emphasis on capabilities; that is, a larger size of project involving various types of work requires different types of capabilities and thus internalisation becomes less likely. The insignificance of ‘structural alteration/addition’ is mainly due to the lack of many respondents’ replies to this open-ended question.

7.3 Likelihood of DIY Renovation: Without Outliers

Three regressions without outliers are reported in this section to see if the removal of outliers would improve the results. After seven, eight and seven outliers were removed from models 2-1, 2-2 and 2-3 in Table 2 respectively, the sample size was reduced to 248 (196 hire-renovators, 52 DIY-renovators), 247 (195 hire-renovators, 52 DIY-renovators) and 247 (197 hire-renovators, 50 DIY-renovators) accordingly.

As seen in Table 3, models 2-1, 2-2 and 2-3 demonstrate higher explanatory powers (i.e. R² values) in the likelihood of DIY renovation.

Table 3.

Dependent variable:0 = hire-renovator	Model 1			Model 2-1			Model 2-2			Model 2-3
Dependent variable:0 = hire-renovator	Fundamental variables only			All the variables			All the variables			All the variables
1 = DIY-renovator				Use (a) ‘time spent on preparation’			Use (a) ‘time spent on preparation’			Use (b) ‘structural alteration/addition’
				Use (c) ‘number of renovation experiences’			Use (d), (e) ‘type of renovation experience’			Use (c) ‘number of renovation experiences’
Independent variables	B	Exp(B)	Wald	B	Exp(B)	Wald	B	Exp(B)	Wald	B	Exp(B)	Wald
Block 1: fundamental variables
Age of house	−0.01	0.99	1.87	0.00	1.00	0.09	−0.02	0.98	3.42*	−0.00	1.00	0.06
Length of residence	−0.03	0.97	1.17	−0.02	0.98	0.32	−0.11	0.90	6.50***	−0.01	0.99	0.11
Age of renovator	−0.03	0.97	3.04*	−0.05	0.95	5.82**	−0.06	0.94	4.69**	−0.05	0.95	5.27**
Level of education	−0.29	0.75	2.82*	−0.42	0.66	3.82**	−0.55	0.58	3.57*	−0.44	0.64	4.20**
Level of household income	−0.16	0.85	3.02*	−0.13	0.88	1.26	0.03	1.03	0.04	−0.15	0.86	1.61
Number of wage-earners	0.64	1.89	6.91***	0.61	1.84	5.16**	1.32	3.73	11.03***	0.47	1.60	2.82*
Time spent in travelling to work	0.00	1.00	0.80	0.00	1.00	0.53	0.00	1.00	0.47	−0.01	0.99	0.92
Block 2: corporate and social/psychological variables
Size of a renovation project
(a) Time spent on preparation	—	—	—	−0.74	0.48	8.73***	−1.13	0.32	11.06***	—	—	—
(b) Structural alteration/addition	—	—	—	—	—	—	—	—	—	−0.19	0.83	0.20
Capability
(c) Number of renovation experiences	—	—	—	0.07	1.08	0.56	—	—	—	0.13	1.13	1.56
(d) Experience in hiring others (yes = 1)	—	—	—	—	—	—	−2.99	0.05	20.97***	—	—	—
(e) Experience in DIY (yes = 1)	—	—	—	—	—	—	4.32	74.85	22.65***	—	—	—
Number of quotes	—	—	—	−0.33	0.72	5.29**	−0.63	0.53	8.67***	−0.61	0.55	14.57***
Level of trust in contractors	—	—	—	−1.44	0.24	17.91***	−1.96	0.14	14.65***	−1.45	0.23	18.65***
Level of social dependence	—	—	—	0.84	2.31	5.47**	0.93	2.55	3.74**	0.37	1.44	1.10
Level of perceived self-performance in renovation	—	—	—	0.18	1.19	0.11	0.01	1.01	0.00	−0.15	0.86	0.08
Level of perceived ease of renovation	—	—	—	1.14	3.12	9.11***	0.28	1.32	0.31	1.02	2.77	7.27***
Tendency to take adventurous holidays	—	—	—	−0.01	0.99	0.00	−0.85	0.43	1.22	0.06	1.06	0.01
Place of birth (Australia = 1)	—	—	—	0.39	1.47	0.51	1.34	3.81	3.44**	0.52	1.67	0.88
(Intercept)	1.54	1.20	1.65	5.67	290.4	10.15	6.41	609.8	6.76	4.65	104.5	7.47
Model evaluation	Model 1			Model 2-1			Model 2-2			Model 2-3
Chi-squared in omnibus test	22.08***			80.56***			146.45***			76.89***
Pseudo R-squared	0.13			0.43			0.70			0.42
The effect of added block	Fundamental variables			Other variables			Other variables			Other variables
Chi-squared in omnibus test	22.08***			58.49***			124.16***			55.44***
Classification rate (percentages)	Model 1			Model 2-1			Model 2-2			Model 2-3
Hire-renovators correctly classified	81.6			85.7			91.3			87.3
DIY-renovators correctly classified	44.3			63.5			80.4			66.0
Overall	73.4			81.0			89.0			83.0

Notes: * p < 0.10; ** p < 0.05; *** p < 0.01.

The pattern of significant variables was similar to the regressions in Table 2. However, another two variables demonstrated predicted positive effects after outliers were removed: ‘social dependence’ (p <0.05) and ‘being born in Australia’ (p <0.05). Renovators who care about others’ opinions are more likely to engage in DIY; renovators born in Australia—a country where people enjoy creating their own things (such as homes) and feel proud of doing so—are more likely to engage in DIY.

8. Conclusion

Williamson’s vertical integration theory has been applied to ‘make or buy’ decisions in the construction industry but not yet to analogous ‘DIY or hiring others’ decisions in home renovation. Thus, this paper presents empirical analyses of renovators’ choice of using their own labour from the perspectives of institutional, socio- and behavioural economics. While demographics including ‘age of renovator’ and ‘number of wage-earners’ are important in explaining renovators’ DIY likelihood, the significant contributions of the contractual factor ‘time spent on preparation’ (based on Williamson’s theory), the capability factors ‘experience in DIY’ and ‘experience in hiring others’ (based on Richardson’s argument) and the psychological factors ‘trust in contractors’ (based on socio- and behavioural economics) support this pluralistic approach.

‘Time spent on preparation’ (a proxy for the size of a renovation project in terms of complexity of tasks) has crucial negative influence on the DIY likelihood. Although this factor, in terms of Williamson’s theory, should have had a positive effect on the DIY likelihood via its impact on the risk of contractual failure, its significant negative effect indicates that renovators are less likely to use their own labour when the size of a project (related to the tasks involved) is too large for them to handle. However, this is consistent with Richardson’s argument in terms of ‘capability’, which is embodied in the type of renovation experience rather than in the number of renovation experiences.

As far as psychological factors are concerned, ‘trust in contractors’ shows its consistent negative influence on the DIY likelihood. Renovators may choose to supply their own labour due to the lack of trust in contractors, or renovators may hire a contractor based on pure trust without conventional cost–benefit analysis. In fact, 12 per cent of surveyed hire-renovators did not ask for any quotes prior to the commencement of renovation; they may have turned to the contractors they had worked with or the contractors recommended by acquaintances. Thus, it may be crucial for further studies to uncover experienced renovators’ rules of thumb in judging contractors’ trustworthiness and for policy-makers or regulatory institutions (such as trade associations in the building sector) to devise ways to help renovators (especially first-time renovators) to find trustworthy contractors. Otherwise, renovators without any prior experience may risk physical harm if their choice of DIY is driven by their lack of trust.

Moreover, ‘perceived ease of renovation’ and ‘social dependence’ variables also show positive influences on DIY likelihood (in the models without outliers). Since renovators’ DIY decisions may be driven by how easy they perceive renovation to be and by how much they are concerned about others’ opinions, it may be important for relevant institutions to raise the awareness of safety in DIY renovation and the difference between the fantasy in DIY renovation shows on TV and the reality in ‘living in and sleeping on’ renovation.

Footnotes

Acknowledgements

The author is grateful to Dr Peter E. Earl and to the anonymous referees for helpful comments on the earlier version of this paper.

Funding Statement

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

Alpers

Gerdes

(2006) Another look at ‘look-alikes’: can judges match belongings with their owners?, Journal of Individual Differences, 27, pp. 38–41.

Arnould

Price

(1993) River magic: extraordinary experience and the extended service encounter, Journal of Consumer Research, 20, pp. 24–45.

Ashby

Ozanne-Smith

Fox

(2007) Investigating the over-representation of older persons in do-it-yourself home maintenance injury and barriers to prevention, Injury Prevention, 13, pp. 328–333.

Bedi

Goldbloom

(2008) A review of nonoccupational ladder-related injuries in Victoria: as easy as falling off a ladder, Journal of Trauma-injury Infection and Critical Care, 64, pp. 1608–1612.

Bikhchandani

Hirshleifer

Welch

(1992) A theory of fads, fashion, custom, and cultural change as informational cascades, The Journal of Political Economy, 100, pp. 992–1026.

Bogdon

(1996) Homeowner renovation and repair: the decision to hire someone else to do the project, Journal of Housing Economics, 5, pp. 323–350.

Brooks

M. A.

Earl

(1987) On the implications of jointness in a normative model of behaviour based on an activity hierarchy, Journal of Consumer Research, 14, pp. 445–448.

Celsi

Rose

Leigh

(1993) An exploration of high-risk leisure consumption through skydiving, Journal of Consumer Research, 20, pp. 1–23.

Chau

K. W.

Walker

(1994) Institutional costs and the nature of subcontracting in the construction industry, in: Rowlinson

S. M.

(Ed.) Proceedings of 13 CIB W92 Symposium: East Meets West: Procurement Systems, pp. 371–377. CIB Publication No. 175. Department of Surveying, University of Hong Kong.

10.

Colvin

(2005) Australia to take 20,000 extra skilled migrants. Australian Broadcasting Corporation (http://www.abc.net.au/pm/content/2005/s1345602.htm; accessed 12 July 2007).

11.

Dunlevy

(2003) Don’t be a building blockhead, The Australian, Sydney, 23 August, p. 33.

12.

Dunning

Johnson

Ehrlinger

Kruger

(2003) Why people fail to recognize their own incompetence, Psychological Science, 12, pp. 83–87.

13.

Earl

(1986) Lifestyle Economics: Consumer Behaviour in a Turbulent World. Brighton: Wheatsheaf.

14.

Earl

(1996) Contracts, coordination and the construction industry, in: Earl

(Ed.) Management, Marketing and the Competitive Process, pp. 149–171. Cheltenham: Edward Elgar.

15.

Earl

(2005) Behavioural economics and the economics of regulation. Briefing paper prepared for the New Zealand Ministry of Economic Development (http://espace.library.uq.edu.au/eserv.php?pid=UQ:8811&dsID=Earl_NZ_MED_repo.pdf; accessed May 22 2008).

16.

Earl

Potts

(2004) The market for preferences, Cambridge Journal of Economics, 28, pp. 619–633.

17.

Earl

Wakeley

(2005) Business Economics: A Contemporary Approach. Maidenhead: McGraw Hill.

18.

Earl

Peng

T.–C.

Potts

(2007) Decision-rule cascades and the dynamics of speculative bubbles, Journal of Economic Psychology, 28, pp. 351–364.

19.

The Economist (2003) Doing up the middle kingdom, 369, 11 October, pp. 67–68.

20.

Ekelund

Mixon

Ressler

(1995) Advertising and information: an empirical study of search, experience and credence goods, Journal of Economic Studies, 22, pp. 33–43.

21.

Finnerty

(2007) Television: house of dreams, The Wall Street Journal, 5 January, p. W.9C.

22.

Garson

(2008) Logistic regression (http://www2.chass.ncsu.edu/garson/PA765/logistic.htm; accessed 5 January 2008).

23.

Granovetter

(1985) Economic action and social structure: the problem of embeddedness, American Journal of Sociology, 91, pp. 481–510.

24.

Granovetter

(2005) The impact of social structure on economic outcomes, Journal of Economic Perspectives, 19, pp. 33–50.

25.

Hoffman

(2006) Home DIY kills, maims and injures, The Australian, Sydney, 4 March, p. 31.

26.

Hooton

(2006) D. I. Why?, The Sydney Morning Herald, 14–16 April, pp. 32–37.

27.

Jamrozik

(1994) Renovation: the pain and the gain, The Independent Monthly, June, pp. 35–41.

28.

Kahneman

Krueger

(2006) Developments in the measurement of subjective well-being, Journal of Economic Perspectives, 20, pp. 3–24.

29.

Kathy

(2007) Home renovations keeps pressure on prices. Australian Broadcasting Corporation (www.abc.net.au/insidebusiness/content/2007/s1940963.htm; accessed 12 June 2007).

30.

Katona

(1960) The Powerful Consumer: Psychological Studies of the American Economy. New York: McGraw-Hill.

31.

Kruger

(2007) Tradies, TAFE unconvinced on training reforms. Australian Broadcasting Corporation (http://www.abc.net.au/worldtoday/content/2007/s1920584.htm; accessed 14 July 2007).

32.

Levitt

Dubner

(2005) Freakonomics: A Rogue Economist Explores the Hidden Side of Everything. Camberwell: Penguin.

33.

Mares

(2008) Tradeswomen are doing it for themselves. Australian Broadcasting Corporation (http://www.abc.net.au/rn/nationalinterest/stories/2008/2314605.htm; accessed 25 August 2008).

34.

Masia

Miller

(2006) Don’t Do-it-yourself Handbook: The Best and Worst of the Weekend Woodies. Sydney: ABC Books.

35.

May

(2005) Planning Your Perfect Home Renovation. Crows Nest, NSW: Allen & Unwin.

36.

Mendelsohn

(1977) Empirical evidence on home improvements, Journal of Urban Economics, 4, pp. 459–468.

37.

Miller

G. A.

(1956) The magical number seven plus or minus two: some limits on our capability for processing information, Psychological Review, 63, pp. 81–97.

38.

Nelson

(1970) Information and consumer behaviour, Journal of Political Economy, 78, pp. 311–329.

39.

Norman

Streiner

(2000) Biostatistics: The Bare Essentials. Hamilton, BC: Decker.

40.

Peng

T.–C.

(2009) A pluralistic analysis of housing renovation choices in Brisbane. PhD thesis, The University of Queensland.

41.

Peng

T.–C.

(2011) Overcapitalization and cost escalation in housing renovation, New Zealand Economic Papers, 45, pp. 119–138.

42.

Phipps

(1983) Housing renovation by recent movers into the core neighbourhoods of Saskatoon, Canadian Geographer, 27, pp. 240–262.

43.

Pollakowski

(1988) The determinants of residential renovation and repair activity. Final report prepared for the Office of Policy Development and Research, US Department of Housing and Urban Development.

44.

Potepan

(1989) Interest rates, income, and home improvement decisions, Journal of Urban Economics, 25, pp. 282–294.

45.

Richardson

(1972) The organisation of industry, The Economic Journal, 82, pp. 883–896.

46.

Routley

Ozanne-Smith

(1995) Prevention of injuries associated with do-it-yourself activities. Accident Research Centre, Monash University (http://www.monash.edu.au/muarc/reports/Other/DHS01.pdf; accessed 2 October 2008).

47.

Scitovsky

(1981) The desire for excitement in modern society, Kyklos, 34, pp. 3–13.

48.

Scitovsky

(1992) The Joyless Economy. New York: Oxford University Press.

49.

Shackle

G. L. S.

(1972) Epistemics and Economics: A Critique of Economic Doctrines. Cambridge: Cambridge University Press.

50.

Simon

(1991) Models of My Life. New York: Basic Books.

51.

Tabachnick

Fidell

(2007) Using Multivariate Statistics. Boston, MA: Allyn & Bacon.

52.

Vandiver

(1972) Why should my conscience bother me?, in: Heilbroner

(Ed.) In the Name of Profit, pp. 3–31. Garden City, NY: Doubleday.

53.

Walker

(2007) Project Management in Construction, 5th edn. Oxford: Blackwell Publishing.

54.

Walker

Chau

K. W.

(1999) The relationship between construction project management theory and transaction cost economics, Engineering Construction and Architectural Management, 6(2), pp. 166–176.

55.

Williamson

O. E.

(1975) Markets and Hierarchies: Analysis and Antitrust Implications. New York: Free Press.

56.

Williamson

O. E.

(1985) The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting. New York: Free Press.

57.

Wyld

(2005) In bed with the builder, The Sydney Morning Herald, 6 October, pp. 8–9.