Other explanatory variables
maids services , In addition to the price variables, we include other measures to control for cross‐household differences in demand for domestic production, taste for domestic production, and ability to pay. Neither dataset contains detailed information on non‐labour income, a variable that may impact ability to pay. For the French data, we construct a dummy variable to identify households that receive rents or dividends. In the case of the UK data, our non‐labour income dummy identifies households receiving rents, dividends or alimony. Dummy variables are also incorporated to identify those living in London or Paris. Numerous factors may differ in these major cities, for example cultural attitudes that influence preferences for domestic work as well as the average size of the residence. A dummy variable to identify the summer season is incorporated in recognition that the housework tasks performed may differ by season, particularly during the summer.

Controls for various household characteristics are also included. One such is a dummy variable to distinguish between married and cohabiting households. Cohabiters may be more independent and invest less in household production because their long‐run opportunity to recoup these investments is more limited. Information on the presence of children of various ages is included primarily with the intent of controlling for differences in demand. The presence of other household adults also suggests a greater need for services. Of course, older children and other adults could help to provide household labour, thus their impact is not clear ex ante.

Dummy variables identifying non‐white persons in the UK and persons not born in France (‘minorities’) are incorporated in recognition that there may be cultural differences in the valuation of home production. Individuals who have a college degree or more are identified, as there exists some evidence that more educated men are more willing to do housework, and that more educated women may be more likely to ‘do gender’ (for a discussion see, for example, Bianchi et al. 2000). Information on the age of each partner is included as age may also capture attitudes towards domestic work.

A complete list of the explanatory variables and their sample statistics is provided in Table 2.

Table 2. Summary Statistics
Variable UK France
Mean S.D. Mean S.D.
Man’s imputed log wage 1.95 0.23 4.08 0.32
Woman’s imputed log wage 1.65 0.25 3.84 0.32
Maid’s log median wagea 1.52 0.04 3.52 0.06
Log of electricity cost per kwh 2.01 0.06
Receives some non‐labour income 0.27 0.44 0.16 0.37
Lives in London/Paris 0.07 0.25 0.02 0.15
Summer 0.25 0.43 0.12 0.33
Cohabits 0.18 0.38 0.21 0.41
Number of other persons age 17+ 0.28 0.60 0.42 0.75
Presence of child aged 0–2 0.17 0.37 0.16 0.37
Presence of child aged 3–4b 0.12 0.32 0.16 0.37
Presence of child aged 5–9b 0.25 0.43 0.28 0.45
Presence of child aged 10–16 0.31 0.46 0.32 0.47
Woman is minority 0.03 0.18 0.06 0.23
Woman has a university degree 0.14 0.34 0.10 0.30
Woman’s age 38.92 9.57 39.30 8.90
Man is minority 0.04 0.20 0.06 0.23
Man has a university degree 0.14 0.34 0.12 0.33
Man’s age 40.92 9.53 41.65 8.93
Number of observations 1295 2924
Notes
Prices are measured in the UK in pounds and in France in euros.
a Reported maid’s wages are gross of taxes and thus represent actual cost. UK wages are net.
b The age ranges for children reported in the table are for the UK. In France we distinguish between children aged 3–5 and children aged 6–9, as children begin attending elementary school at age 6 in France and age 5 in the UK.
3 Results
The estimated common price effects are presented in Table 3, as are the effects of non‐labour income. Panel A reports the results for the UK, and Panel B those for France. Appendix Table A1 presents the other covariate results for the UK, and Appendix Table A1 does the same for France.

Table 3. Price Effects
Explanatory variables His imputed log wage Her imputed log wage Maid’s log median wage Non‐labour income dummy
Dependent variable Coeff. Coeff. Coeff. Coeff.
Panel A: UK
Hire a maid 1.0586** 1.3172* −2.4595 0.2388* (0.4287) (0.4514) (2.3851) (0.1331) [0.0845] [0.1052] [−0.1963] [0.0191] Have a dishwasher 1.9081* 0.4912** −0.0446 0.0560
(0.2767) (0.2478) (1.4453) (0.0854)
[0.7570] [0.1949] [−0.0177] [0.0222]
His housework time
Weekdays −2.1492 3.2241 −3.4898 3.2893
(5.7658) (6.4685) (28.4262) (2.1541)
Weekends −7.4829 21.5920* 108.2303 1.7715
(8.7900) (8.2605) (47.9785) (3.1076)
Her housework time
Weekdays −11.1301 −30.0979** −13.2681 1.1082
(13.8940) (15.2429) (79.4209) (5.0913)
Weekends −25.4533 7.8362 96.1676 2.7097
(15.9750) (14.3591) (83.8524) (5.3491)
Panel B: France
Hire a maid 0.5230** 1.2644* −2.0090 0.3958*
(0.2657) (0.2235) (1.5023) (0.0976)
[0.0262] [0.0632] [−0.1004] [0.0198]
Have a dishwasher 1.0504* 0.8099* −2.3024** 0.4177* (0.1957) (0.1329) (0.9289) (0.0793) [0.3969] [0.3061] [−0.8700] [0.1578] His housework time Weekdays −1.8379 7.0443 1.8407 0.4241
(3.7490) (2.8753) (21.3736) (1.8415)
Weekends −1.0203 20.9187* 97.1586 −1.8433 (9.8166) (6.8564) (60.9713) (4.9848) Her housework time Weekdays 23.0708* −73.1494* −61.0661 −2.7809
(12.0264) (8.0778) (71.5604) (4.1093)
Weekends 3.4796 −37.0082** 74.5009 −21.0213**
(19.3589) (15.1355) (107.0462) (8.4429)
Notes
Standard errors in parentheses. Marginal effect in square brackets.
***, **, * indicate statistical significance at the 1%, 5%, 10% level, respectively, for two‐tailed tests.
Also included in the specification are dummy variables to identify those living in London (Paris), those cohabiting, the presence of children of various ages, individuals who are in poor health and a summer interview; a continuous measure of the number of other adults in the household; and the age, university degree status and minority status for each partner.
Marginal effects are calculated for married couples with sample mean opportunity costs, sample mean maids services and electricity log prices, who are of approximately sample mean age and have a teenager. All other covariates are zero.