Document View

The factors that influence the income of real estate licensees are examined. An empirical human capital earnings model is developed from a 1995 survey of Florida real estate brokers and salespeople. In seeking to explain earnings of real estate licensees, several additional human capital components are measured. A number of factors are seen to positively affect licensee income. These include: 1. number of hours worked, 2. experience, 3. franchise affiliation, 4. being an owner/manager, 5. working in a metropolitan area, 6. level of job satisfaction, and 7. having errors and omissions insurance. Variables that have a negative effect on income include: 1. being a female, 2. selling primarily residential properties, 3. age of licensee, 4. image perception, and 5. working weekends. Segmenting the data by income into thirds and comparing the means of the variables for the high- and low-income groups, several variable means are significantly different.

Introduction

Real estate sales has been characterized as an industry with high turnover, low per capita income, and increasing competition (see Johnson, Dotson, and Dunlap, 1988). Competing in this type of industry requires a critical understanding of those factors that maximize customer satisfaction and improve service quality. The principal asset of the real estate sales firm is its workforce; understanding this human capital component is thus critical to the success of the firm. Prior empirical research of human capital theory has sought to explain differences in earnings across industries based upon characteristics of the occupation or the human capital used in a given industry (see, for example, Polachek, 1981; McDowell, 1982, and Willis, 1986).

Several previous studies have examined the determinants of earnings of real estate licensees (see Follain, Lutes and Meier, 1987; Crellin, Frew and Jud, 1988; and Glower and Hendershott, 1988). All these studies have utilized a human capital model similar to the one employed in this study to explain earnings of real estate licensees. The purpose of this work is to further examine the determinants of real estate licensee income by examining additional human capital components.' The data are results from a 1995 survey of a random sample of Florida real estate licensees that includes both brokers and salespeople.

Determining Licensee Income

The standard human capital earnings model sets income as a function of education and experience (see Mincer, 1974). Earnings, of course, are considered to be positively related to education at a constant rate. The rationale is that each licensee begins his/her career with a stock of human capital determined by the level of education. Additional increases in this level of human capital would presumably be constant across licensees since all licensees face the same continuing education requirements.2 Experience is typically measured as a quadratic function with experience being positively related to income and a squared term to measure nonconstant marginal returns. Rosen (1976) significantly developed this aspect of the human capital earnings model by incorporating experience as part of a lifetime earnings function.

In addition to experience and education, other variables have been included in previous studies to measure the human component of real estate sales (see Follain et al., 1987; Crellin et al., 1988; Glower and Hendershott, 1988). Such variables have been the number of hours worked, type of license, location, gender, race, age, etc. Equations have also typically included some firm characteristics such as size of firm and whether the firm is a franchise affiliation.

This study expands the human capital aspect of the real estate sales profession by examining additional licensee characteristics that may have some effect on the earnings of a licensee. This allows certain interesting questions regarding licensee income to be answered. For example, once a relationship is established, does a buyer's or seller's loyalty lie with the realty firm or the salesperson? In other words, do those salespeople who change affiliations earn less than those who remain with one firm? In addition, do those licensees who choose real estate sales as a second or third occupation earn less than those who choose real estate as their first occupation? Also, is willingness to work weekends a factor in a licensee's earnings?

Presumably, social or professional interaction is important in a profession such as real estate sales so as to maintain or make "contacts". Does belonging to clubs or professional organizations such as the Chamber of Commerce or the Country/Golf Club have an effect on earnings?3

Does a real estate licensee's attitude toward the profession have a bearing on his/her income? The survey asked respondents to describe the image of the real estate agent in the public's eye as professional, salesperson, necessary evil, or a huckster. This study examines whether those licensees who feel they are viewed as professionals earn more than other licensees. Also, respondents were asked to give their general level of satisfaction with the profession. This would allow a determination of whether the level of job satisfaction affects income.

The real estate sales industry is currently sensitive to certain legal problems. These include dual agency, disclosure, etc. The study examines whether having errors and omissions insurance affects licensee income. It also contrasts the income of licensees who received some compensation from the seller to the income for licensees who derived their income strictly from the buyer.

Education is assumed to have a positive effect on earnings (Becker, 1975). Some licensees have more general education than others (such as some college education versus merely finishing high school). For real estate-specific education within a single state, most licensees' education in terms of pre-licensing hour requirements would be comparable. However, does the source of those pre-licensing hours affect earnings? Survey results showed that 72% of respondents completed their original pre-licensing hours requirement through private real estate schools while 22% obtained their pre-licensing hours through a college or university. Also, as post-licensing continuing education requirements are fulfilled, does it matter whether those hours are obtained in a classroom setting or by correspondence? Thirty-two percent of survey respondents indicated that they completed none of their continuing education hours through correspondence while 52% indicated that 76o/-100% of their continuing education hours were completed through correspondence.

Typical firm characteristics considered to affect licensee earnings appearing in previous studies are firm size and franchise affiliation. This study will also examine whether being a member of a national referral network or having a relocation/management company affects licensee income.

The Data and Empirical Model

The Model

Following the human capital approach of previous studies, the empirical model takes the form:

INCOME = f(WORK PROFILE, EDUCATION, EXPERIENCE, PERSONAL CHARACTERISTICS, PROFESSIONAL CHARACTERISTICS, FIRM CHARACTERISTICS, LOCATION),

where INCOME is the licensee's personal income from real estate activities for 1994. The model uses the natural log of income as the dependent variable.4 The variables contained in the remaining categories are described below. Exhibit I contains variable definitions and Exhibit 2 provides summary statistics for the variables included in the model. Exhibit 2 shows that the average annual income per licensee was $37,380.5 About one-fourth of licensees had 1994 annual income of $10,000 or less and about three-fourths of licensees earned $50,000 or less for 1994.

Work Profile. This category contains the natural log of the number of hours per week spent working in real estate activities (In HRS). A positive relationship is expected between the number of hours worked and earnings. The average number of hours worked per week was 36.18.6 The binary variable FTIME indicates that about 72% of licensees worked at least twenty-eight hours per week. The equation also contains a variable measuring the average number of weekend days worked per month (WKEND) which had a mean of 3.263 days. The relationship between WKEND and income is unclear. On one hand, a greater expenditure of effort should generate greater earnings. On the other hand, a licensee with lower income may be willing to work weekends. In this case, a negative sign would be observed on this variable. Moreover, weekend work is generally associated with residential real estate and the literature has shown that those licensees selling primarily residential real estate earn less, on average.

Education. Education variables in the model include: EDYRS, which measures the number of years of formal education; the source of pre-licensing education denoted by PRELIC; and CORR, which indicates whether or not continuing education requirements were completed by correspondence. Since more education would increase the licensee's base of human capital, EDYRS is expected to have a positive relationship with earnings. Exhibit 2 shows that about 83% of respondents had some education past high school.7

Enlarge 200% Enlarge 400%

Enlarge 200% Enlarge 400%

About 72% of respondents indicated that they fulfilled their pre-licensing education requirements through private real estate schools. As Exhibit 2 shows, about 20% indicated a college or university as their source of pre-licensing education. Small percentages of respondents indicated sources such as vocational/technical schools, the Appraisal Institute, etc. This variable allows an exploration of whether the marginal contribution to human capital is different across education sources. Also, about 68% of respondents indicated that they satisfied at least some of their continuing education hours through correspondence.

Experience. The experience variables include the number of years the licensee has been actively engaged in the real estate business (EXP) and an experience squared term (EXPSQ) to capture any nonlinear marginal returns to experience. The relationship between income and experience is expected to be positive. A negative coefficient for experience squared would indicate decreasing marginal returns to experience. Exhibit 2 shows that the average licensee had about thirteen years of experience.8

Personal Characteristics. Personal licensee characteristics considered include age (AGE), gender (GEN), and minority race (MIN). As seen in Exhibit 2, the average age of licensees was about fifty years and about 56% were males.9 Only about 8% of respondents indicated being a minority. Also included is a social interaction variable which should have a positive effect on income. This variable is measured by membership in a club or professional organization (CLUB). This variable has a value of one if the licensee indicated belonging to any of a number of clubs or professional organizations.'o These included the Chamber of Commerce, Country/Golf Club, Kiwanis, Elks, Women's Club, etc. Exhibit 2 shows that about 66% of respondents reported belonging to at least one club or organization.

Presumably, a licensee's behavior would reflect the licensee's understanding of public perception. To measure this, a variable is included to reflect perceived image (IMAGE). This variable has a value of one if the licensee feels he/she is viewed as a professional and zero otherwise. " About 27% of respondents felt they were viewed as professionals. Also, assuming that the licensee's work habits reflect his/her level of job satisfaction, a variable (SATIS) is included that has a value of one if the licensee is completely or generally satisfied with the profession (about 57% of respondents) and zero otherwise.

Professional Characteristics. Professional characteristics are examined to determine their effect on income. These include the type of license held (BROK, which is one if the licensee holds a broker's license (about 43% of respondents) and zero if a salesperson's license is held). The variable CURR provides a measure of the value of stability in generating income by measuring how long the licensee has been affiliated with his/her current firm. The average is 6.26 years.

To determine whether some activities in which real estate professionals engage produce greater income than others, two variables are included in the model. The variable RES is included to indicate the activity to which the licensee devotes most of his/her time. This variable has a value of one if the primary activity is residential property sales and zero otherwise. About 69% of respondents indicated that selling new/existing residential property was their major focus. Also, to determine income differences for typical licensees versus those who are firm owners or firm managers, MGTOWN is included which has a value of one if the licensee is an owner and/or manager (about 26% of respondents) and zero otherwise.

To determine whether the degree of professionalism attained by the licensee has an effect on income, the variable DESIG is included. DESIG has a value of one if the licensee holds a professional designation and zero otherwise. About 27% of licensees reported holding a professional designation. The GRI was the major designation reported with about 16% of respondents holding this designation. Respondents indicated the holding of designations such as the CRS, CRB, etc. One would expect professionalism to have a positive impact upon earnings since holding the designation indicates a certain amount of effort and discipline on the part of the licensee.

In order to measure any advantage in being a member of an association that may allow the licensee to stay better informed of changes in the real estate environment and, in general, to stay closer to the market, a variable (FAR) is included to indicate whether or not the licensee is a member of the Florida Association of Realtors. About 78% of licensees indicated membership in FAR.

To determine the effect of agency form on income the variable SELLER is included. This variable allows a comparison of income between those licensees who derived some or all their income from the seller as opposed to those licensees whose total income came from the buyer. In the former case, the presumption is that licensees were operating under the common cooperative relationship in the form of subagency while in the latter case, where no income is from the seller, presumably the "buyer's brokerage" form of agency existed.12 The survey results show that about 80% of licensees reported deriving at least some income from the seller.

AFFIL is included to determine whether mobility has an effect on licensee earnings. This variable, which is the number of real estate firms with which the licensee has been affiliated, should measure whether more frequent changes in affiliation affect earnings. The effect on income is unclear. On one hand, a licensee may change affiliations to maximize commission splits and therefore increase income. On the other hand, some customer income could be lost due to frequent affiliation changes.

Historically, Florida has experienced substantial in-migration of residents. Some people, having retired from one occupation, may choose real estate sales as a second, third, etc. career. CAREER measures whether real estate is the licensee's first, second, third, or fourth (or more) occupation.

Finally, on a professional level, the variable E&O indicates whether or not the licensee has errors and omissions insurance. About 58% of respondents indicated they held E&O insurance which was paid for either by the licensee, the firm, or both. This variable should be positively related to income since having this protection may be an incentive for the licensee to be more aggressive in pursuing sales.

Firm Characteristics. Does the type of business with which the licensee is affiliated have an effect on individual licensee earnings? Can licensees be more successful with one type of firm versus another? To examine these questions, the variable FRAN is included which distinguishes franchise firms from independent, stand-alone firms. About 30% of respondents indicated they worked with a national franchise company. Also, the variable REFREL indicates whether the real estate firm belongs to a national referral network and/or is a relocation/management company (25% of respondents). Lastly, the variable SIZE indicates the size of the real estate firm in terms of the number of licensee affiliates. Exhibit 2 shows that the average firm had thirty-six licensee affiliations.13

Location. A location variable is included to determine whether those licensees working in major metropolitan areas earn more than those working in lesser populated areas. The variable METRO has a value of one if the licensee indicated working in a major metro area (about 29% of respondents) and zero otherwise. In general, one would expect a positive relationship with earnings since licensees in more densely populated areas should have greater opportunities to generate income.

The Data

The data are the results of a 1995 survey sent to a random sample of 1200 active real estate brokers and salespeople in Florida. Of the respondents, a total of 185 observations containing complete data are used in the study. The characteristics measured in the survey included (1) personal information such as age, experience, income, education, etc., (2) sources of education (both pre- and post-licensing), (3) professionalism (holding of professional designations, membership in the Florida Association of Realtors, etc.), (4) firm characteristics such as size, franchise affiliation, etc.

Results

The results for the empirical models are given in Exhibits 3 and 4.

The Full Model

Exhibit 3 contains the results for the full regression model. The model has an adjusted R2 of 61% and the variables behave generally as expected.'4

Work Profile. Not surprisingly, the number of hours worked has the strongest effect on income.15 However, the FTIME variable indicates that those licensees who work full-time do not earn more per hour than those who work less hours. Although WKEND is not significant at the 10% level, the size of the t-statistic and the negative coefficient imply that those licensees who work more weekend days tend to earn less.

Education. The lack of significance for the education variable (EDYRs seems to indicate that the level of education does not create differences in human capital that may significantly affect income.16 Likewise, the source of pre-licensing education (primarily private real estate schools versus colleges and/or universities) has no significant effect on income. Lastly, there is no significant difference in earnings for those licensees who use correspondence to satisfy continuing education requirements versus those who do not.

Experience. Both experience variables are significant. The positive coefficient for EXP indicates that income increases with experience. However, the negative coefficient on EXPSQ shows decreasing marginal returns to experience. This result is consistent with Glower and Hendershott (1988) who also show that experience increases the productivity of licensees but, beyond some point, additional experience is of lesser value.

Personal Characteristics. Several personal characteristics of licensees affect income. Age is negatively related to income while the gender variable, GEN, shows that income for male licensees is significantly higher than income for female licensees. Neither the minority (MIN) nor the club (CLUB) variable is significant. The negative coefficient for IMAGE is interesting. Recall that this variable indicates that the licensee feels that the market perceives him/her as a professional. This result implies that licensees who feel they are viewed as professionals earn less than those who feel they are viewed by the public as salespeople, hucksters, or necessary evils.17 The significant positive coefficient for SATIS shows that the more satisfied the licensee is with the profession, the greater the income.18

Professional Characteristics. The results show no significant difference in income between brokers and salespeople.19 Also, the length of time a licensee is affiliated with one firm does not significantly affect income. Those licensees who primarily sell residential property earn less than others. Being an owner of a real estate firm or being involved in firm management helps licensees earn greater income.

Enlarge 200% Enlarge 400%

The lack of significance for DESIG seems to indicate that the holding of professional designations does not contribute to a higher base of human capital which may increase income. Nor does there seem to be an advantage in belonging to the Florida Association of Realtors (FAR).

The lack of significance for the SELLER variable, which measures the source of income (seller versus buyer), indicates that licensee income derived from both the seller and buyer is not significantly different from income derived exclusively from the buyer. The results do show that those licensees who carry errors and omissions insurance (E&O) earn significantly higher incomes than those who do not.

The AFFIL variable shows that income is not affected by the number of affiliations that the licensee has had. It appears there is no advantage to either maintaining a constant affiliation or to more frequent changes. Also, the CAREER variable shows that licensees entering the real estate profession as a first occupation do not earn significantly higher incomes than those who choose real estate as a second, third, etc. career.

Firm Characteristics. Only one firm characteristic is significant. The FRAN variable shows that licensees affiliated with franchise firms have higher incomes than other licensees. This could be due to greater name recognition or other factors such as 100% commission firms. However, neither the size of the firm nor being a member of a referral or relocation service has a significant effect on income.

Location. The significant coefficient for METRO indicates that licensees working in major metropolitan areas generate significantly higher average income than their counterparts in less populated areas. This could be the result of a number of factors including higher real estate prices and a greater number of transactions.

The Final Model

In order to better compare these results to previous studies, Exhibit 4 gives the results for a reduced model comparable to models from previous studies. These results are consistent with those in Exhibit 3 with the exception that the WKEND variable becomes significant. Also, the explanatory power of the model is increased as indicated by the higher R2 of 63%.

Comparisons across Studies Variables Common across Studies

Exhibit 5 provides a comparison of the results from this study with previous studies. As seen, the studies are in general agreement on most factors. For example, all studies show a positive effect on income of hours worked and experience. At least three studies show a positive effect of (a) type of license, (b) schooling, (c) being an owner/manager, and (d) working in a metro area; and a negative effect of (a) selling residential property and (b) being a female. Other common results across studies are: (a) a negative effect for experience squared and (b) a positive effect for professional training.

Enlarge 200% Enlarge 400%

The Sirmans and Swicegood (S&S) study and the Crellin, Frew and Jud (CF&J) study (1988) examine the effect of race and age on income. While CF&J find that minorities earned less, S&S find no significant effect of race on income. In contrast, while S&S find a negative effect of age on income, CF&J find no significant relationship between age and income.20 While Follain, Lutes and Meier (FL&M) (1987) and CF&J (1988) find that income increased with firm size, S&S find no significance between firm size and income.21

Conflicting results are found in S&S and CF&J for franchise affiliation. S&S find that franchise affiliation increases income while CF&J find the opposite.

Enlarge 200% Enlarge 400%

Variables Unique to the Sirmans and Swicegood Study

This study seeks to expand the analysis of the determination of real estate licensee income by examining additional factors that may contribute to the licensee's base of human capital. Some of these factors have a significant effect on income. For example, licensees who work on weekends earn less. Job satisfaction is seen to have a positive effect on income and those licensees who perceive themselves as being viewed by the public as professionals earn less. Also, licensees holding errors and omissions insurance earn more than those who do not.

Also interesting are the variables that have no significant effect on income. These include (a) source of pre-licensing education, (b) use of correspondence for continuing education, (c) membership in clubs and/or professional organizations, (d) years with current firm, (e) professional designations, (f) membership in FAR, (g) source of income (buyer versus seller), (h) access to a referral/relocation network, (i) the number of affiliations the licensee has had, and (j) whether the real estate profession is the licensee's first, second, etc. occupation. Though they may often be assumed to benefit the licensees in their profession, these variables in fact appear to add no direct economic advantage.

Testing for Differences in Means

For further insight, the sample is segmented into thirds by income, and characteristics of the top one-third income producers are compared to the bottom one-third by testing for differences in the variable means. The results are given in Exhibit 6. The differences in the variable means, shown in column two, are calculated by subtracting the variable means for the lower income segment from the variable means for the upper income segment.

Results indicate the top income producers have a different work profile than lower income earners. For example, they earn significantly more per hour than the licensees in the bottom one-third. Higher income earners also work significantly more hours per week and more of them work full-time. They also work more on the weekends than lower income licensees.

There does not seem to be a significant difference in education across the two income segments. There is no significant difference in the means of the number of years of formal education nor in the source of pre-licensing education. However, a greater percentage of higher income earners utilized correspondence to satisfy continuing education requirements. There is a difference in the means for experience between the two groups with the higher income group having nearly six more years of experience.

There is very little difference in the personal characteristics across the two income groups. There is no significant difference in the means for age, gender, minority, image, or level of satisfaction. The CLUB variable does show that a greater percentage of higher income licensees belong to clubs and/or professional organizations.

For professional characteristics, the means for several variables are significantly different. A greater proportion of higher income licensees (a) hold a broker's license, (b) have worked longer at their current firm, (c) are involved in ownership/management, (d) hold professional designations, and (e) belong to the Florida Association of Realtors.

A greater percentage of lower income licensees sell residential property as their major focus. There was no significant difference in the means across the two income groups for (a) their source of income (buyer versus seller), (b) having errors and omissions insurance, (c) the number of affiliations over their real estate career, and (d) the number of occupation (first, second, etc. career).

Enlarge 200% Enlarge 400%

The real estate firms with which higher income licensees are affiliated do not appear to have significantly different characteristics from the firms for which lower income licensees work. None of the firm characteristics have significantly different means across the two income groups.

The metro variable indicates that, relative to lower income earners, higher income licensees are not clustered in metropolitan areas. There is no significant difference in the means across the two income groups.

Summary

This study has examined the factors that influence the income of real estate licensees. An empirical human capital earnings model was developed from a 1995 survey of Florida real estate brokers and salespeople. In seeking to explain earnings of real estate licensees, this study has expanded previous ones by measuring several additional human capital components.

A number of factors were seen to positively affect licensee income. These include (a) number of hours worked, (b) work experience, (c) franchise affiliation, (d) being an owner/manager, (e) working in a metropolitan area, (f) level of job satisfaction, and (g) having errors and omissions insurance. Variables that had a negative effect on income include (a) being a female, (b) selling primarily residential properties, (c) age of licensee, (d) image perception, and (e) working weekends.

Human capital components that did not significantly affect income were (a) type of license, (b) schooling, (c) race, (d) size of firm, (e) source of pre-licensing education, (f) use of correspondence in continuing education, (g) membership in various clubs or professional organizations, (h) length of affiliation with current firm, (i) professional designations, (j) membership in the state's Realtor association, (k) receiving income from the seller versus the buyer, and (1) having access to a referral/relocation network.

When the data were segmented by income into thirds and the means of the variables for the high- and low-income groups were compared, several variable means were significantly different. The high-income group had significantly higher means for these variables: (a) hourly income, (b) number of hours worked, (c) working full-time, (d) working on the weekend, (e) utilizing correspondence to satisfy continuing educational requirements, (f) work experience, (g) membership in clubs/professional organizations, (h) holding a broker's license, (i) length of current affiliation, (j) being a manager/owner, (k) holding professional designations, and (1) belonging to the state's Realtor association. The low-income group had a significantly higher variable mean for participation in residential property sales.

The variables for which means were not significantly different across the two income groups were: (a) number of years of education, (b) source of pre-licensing education, (c) age, (d) gender, (e) minority, (f) image, (g) level of satisfaction, (h) source of income (seller versus buyer), (i) having errors and omissions insurance, (j) number of firm affiliations the licensee has had, (k) whether real estate is the licensee's initial occupation, (1) whether the firm is franchise affiliated, (m) whether the firm is a member of a referral network, (n) size of real estate firm, and (o) whether the licensee works in a major metropolitan area.

[Author Affiliation]

*Department of Insurance, Real Estate and Business Law, College of Business, Florida State University, Tallahassee, Florida 32306-1042.

**Department of Finance, College of Business, Florida State University, Tallahassee, Florida 32306-1042. Date Revised-October 1996; Accepted-November 1996.