Thursday, May 04, 2006
Who Wears the Pants? : Bargaining Power in the Housing Market
Motivation:
I want to investigate the market for single family homes. In particular, I am interested in determining how a house or apartment can be designed (through the use of space) most profitably—maximizing value while holding cost relatively constant. While many hedonic price regressions have revealed the value-added of particular amenities, I will examine the differential value-added between “male” and “female” amenities.
The underlying question here is: Who has the stronger bargaining position when it comes to selecting a home, the man or the woman? One piece of anecdotal evidence suggests that women have the upper hand when choosing a home. According to a real estate investor whom I spoke with, women’s amenities disproportionately raise the price of the home compared to their cost. His explanation was that women have stronger preferences regarding homes than men. He believed that amenities such as a high-end kitchen, dining rooms and laundry facilities were much more important to women than rooms such as a basement or an office are to men. As such, men who wish to exert effort to select the home will have to incur large costs in argument with a woman (the same goes for women). However, because women, on average, derive greater benefits from their set of preferred amenities they will be more willing to compete for bargaining power.
Empirical Strategy:
This paper will use two key steps to analyze the hypothesis:
(1) I will determine differences in demand across gender. I will compare the homes of only single men and single women to isolate the gender differences. This method will eliminate the demand of a partner. However, the flaw in this approach is that singles (both men and women) may demand different amenities when they are with a partner. Unfortunately, this problem seems unavoidable.
(2) Based on the findings on differences in demand across gender in part (1), I will test if homes with amenities demanded more strongly by women are more valuable (holding constant other major characteristics). In this part, I will compare only the homes of married couples. By using this method, the results will help explain the dynamics of intra-relationship bargaining in selecting a home.
Empirical predictions:
In part (1), I predict that amenities such as fully-functioning kitchens, dining rooms, and laundry facilities are more strongly demanded by women. Consequently, I expect to find these amenities to be more prevalent in the homes of single women. In addition, I predict that amenities such as basements/attics and office spaces are more strongly demanded by men.
In part (2), I predict that, holding other major housing characteristics constant, female amenities increase home value, while male amenities have no appreciable effect.
If find these hypotheses correct, then I will conclude that women have stronger bargaining power in housing decisions and that designing homes catered to women will, on average, reward the investor with greater returns.
Regressions: I will discuss this in the Results section for now.
Data:
Description:
“The American Housing Survey is conducted by the Bureau of the Census for the Department of Housing and Urban Development (HUD).
The American Housing Survey (AHS) collects data on the Nation's housing, including apartments, single-family homes, mobile homes, vacant housing units, household characteristics, income, housing and neighborhood quality, housing costs, equipment and fuels, size of housing unit, and recent movers. National data are collected in odd numbered years, and data for each of 47 selected Metropolitan Areas are collected currently about every six years. The national sample covers an average 55,000 housing units. Each metropolitan area sample covers 4,100 or more housing units.
Click here to see preliminary results.
I want to investigate the market for single family homes. In particular, I am interested in determining how a house or apartment can be designed (through the use of space) most profitably—maximizing value while holding cost relatively constant. While many hedonic price regressions have revealed the value-added of particular amenities, I will examine the differential value-added between “male” and “female” amenities.
The underlying question here is: Who has the stronger bargaining position when it comes to selecting a home, the man or the woman? One piece of anecdotal evidence suggests that women have the upper hand when choosing a home. According to a real estate investor whom I spoke with, women’s amenities disproportionately raise the price of the home compared to their cost. His explanation was that women have stronger preferences regarding homes than men. He believed that amenities such as a high-end kitchen, dining rooms and laundry facilities were much more important to women than rooms such as a basement or an office are to men. As such, men who wish to exert effort to select the home will have to incur large costs in argument with a woman (the same goes for women). However, because women, on average, derive greater benefits from their set of preferred amenities they will be more willing to compete for bargaining power.
Empirical Strategy:
This paper will use two key steps to analyze the hypothesis:
(1) I will determine differences in demand across gender. I will compare the homes of only single men and single women to isolate the gender differences. This method will eliminate the demand of a partner. However, the flaw in this approach is that singles (both men and women) may demand different amenities when they are with a partner. Unfortunately, this problem seems unavoidable.
(2) Based on the findings on differences in demand across gender in part (1), I will test if homes with amenities demanded more strongly by women are more valuable (holding constant other major characteristics). In this part, I will compare only the homes of married couples. By using this method, the results will help explain the dynamics of intra-relationship bargaining in selecting a home.
Empirical predictions:
In part (1), I predict that amenities such as fully-functioning kitchens, dining rooms, and laundry facilities are more strongly demanded by women. Consequently, I expect to find these amenities to be more prevalent in the homes of single women. In addition, I predict that amenities such as basements/attics and office spaces are more strongly demanded by men.
In part (2), I predict that, holding other major housing characteristics constant, female amenities increase home value, while male amenities have no appreciable effect.
If find these hypotheses correct, then I will conclude that women have stronger bargaining power in housing decisions and that designing homes catered to women will, on average, reward the investor with greater returns.
Regressions: I will discuss this in the Results section for now.
Data:
Description:
“The American Housing Survey is conducted by the Bureau of the Census for the Department of Housing and Urban Development (HUD).
The American Housing Survey (AHS) collects data on the Nation's housing, including apartments, single-family homes, mobile homes, vacant housing units, household characteristics, income, housing and neighborhood quality, housing costs, equipment and fuels, size of housing unit, and recent movers. National data are collected in odd numbered years, and data for each of 47 selected Metropolitan Areas are collected currently about every six years. The national sample covers an average 55,000 housing units. Each metropolitan area sample covers 4,100 or more housing units.
Click here to see preliminary results.
// posted by BW @ 10:56 PM 
Comments:
<< Home
Andy,
In the first part, the important controls are individual controls, not house controls. That is, you want to control for age, race, education, urbanicity, location, etc.. Gender is exogenous, but among the subset of people who live alone gender may be correlated with these variables. And if they are correlated with amenity demands, these regressions may have OVB.
Is the first set of regressions from the metro or national data?
Be sure to include in your paper some explanation of why men and women have different housing preferences.
There seems to be something wrong with your second regressions. First, you only have 112 observations. This is likely why you aren't getting strong results. One of the variables must be missing most of the observations.
Second, when running these types of regressions it is typically to use the log of value as the dependent variable (because of the way the distribution looks and because the coefficient is interpretable as a percent change). So gen lnvalue=ln(value) and use that as the dependant variable.
Third, you should include all of amenities you are interested in the same regression and include municipality fixed effects. E.g.,
areg lnvalue [amenitities] [controls], absorb(town) cl(town)
(The last piece of code clusters your standard errors, which you should do in this case.)
Also, if you can, you should try and add controls for neighborhood characteristics and the distance to the city center. (These may be correlated with the presence of HH amenities and with value, and thus omitting them leads to OVB.)
Finally, to get more at the bargaining issue, maybe we want to see if within couples we can predict what kind of amenities the couple has. What info is available on each of the members of the couple?
Post a Comment
In the first part, the important controls are individual controls, not house controls. That is, you want to control for age, race, education, urbanicity, location, etc.. Gender is exogenous, but among the subset of people who live alone gender may be correlated with these variables. And if they are correlated with amenity demands, these regressions may have OVB.
Is the first set of regressions from the metro or national data?
Be sure to include in your paper some explanation of why men and women have different housing preferences.
There seems to be something wrong with your second regressions. First, you only have 112 observations. This is likely why you aren't getting strong results. One of the variables must be missing most of the observations.
Second, when running these types of regressions it is typically to use the log of value as the dependent variable (because of the way the distribution looks and because the coefficient is interpretable as a percent change). So gen lnvalue=ln(value) and use that as the dependant variable.
Third, you should include all of amenities you are interested in the same regression and include municipality fixed effects. E.g.,
areg lnvalue [amenitities] [controls], absorb(town) cl(town)
(The last piece of code clusters your standard errors, which you should do in this case.)
Also, if you can, you should try and add controls for neighborhood characteristics and the distance to the city center. (These may be correlated with the presence of HH amenities and with value, and thus omitting them leads to OVB.)
Finally, to get more at the bargaining issue, maybe we want to see if within couples we can predict what kind of amenities the couple has. What info is available on each of the members of the couple?
Subscribe to Post Comments [Atom]
<< Home
Subscribe to Posts [Atom]