Document View

Using a sample of newly licensed REALTOR Associates(R), survival rates are calculated for (1) licensees that maintained active status in the Board and (2) licensees that stayed with their original broker. No persistent significant differences in survival rates based on licensee gender, firm size, or franchise affiliation are discovered. Survival rates are significantly related to several measures including total licensees per capita. Tests conducted on survey data identify numerous variables that differed between licensee groups; several of which could be used as a screening device. The results also indicate that brokers can reduce turnover by ensuring that new licensees have accurate expectations. [PUBLICATION ABSTRACT]

Introduction

The high turnover rate for newly licensed real estate agents is a major concern of the brokerage industry. High turnover is problematic for both new licensees and the brokers employing them. New licensees may encounter substantial costs in obtaining their licenses; money that likely would have been better spent by those whose experience in the brokerage industry is short and unsuccessful.1 From the employing broker's perspective, there are both direct and indirect costs associated with high turnover. While indirect costs, such as lost sales and office inefficiency, are difficult to measure, direct costs are measurable and can be significant. For example, high turnover rates result in extra training costs that may be reduced with lower agent turnover.2

Despite practitioner concern about new licensee turnover, little research has been . devoted to this subject. The purpose of this study is to help fill this gap in the literature by examining new licensee survival from two perspectives. From the broker's perspective, licensees that stayed with their original broker are compared to licensees that did not. Comparing these groups may enable brokers to increase the probability that new licensees are successful and remain with the firm. From the industry perspective, licensees that remained in the local Board are compared to licensees that exited the Board.3

Two methodologies are employed. In the first, survival rates are calculated using a sample of all licensees new to the Dayton Board of REALTORS, the Stark County Association of REALTORS and the Mansfield Board of REALTORS during the years 1996 through 1998. With survival defined as the licensee remaining active in the local Board, only 47.5% of licensees survived to the end of their third calendar year. With survival defined as remaining with the original broker, the three-year survival rate drops to 38.5%. No persistent significant differences in survival rates based on licensee gender, firm size, or franchise affiliation are discovered. However, survival rates in the Stark County Association were significantly higher than the other two Boards and this appears due, at least in part, to the fact that there are significantly fewer total licensees per capita in the Stark County Association market area.

The second methodology involves the analysis of survey information obtained from 147 licensees new to the Dayton Board of REALTORS, the Stark County Association of REALTORS, or the Mansfield Board of REALTORS(R) during the years 1996 through 1998. Twenty variables are identified that differ significantly between licensees that remained in the local Board and those that did not, and between licensees that stayed with their original broker and those that did not. Brokers could use several of these variables as screening devices when interviewing prospective sales associates. The results also indicate that brokers can reduce new licensee turnover by ensuring that new licensees have accurate expectations about their duties.

The remainder of the paper is organized as follows. In the next section, a brief review of related literature is presented. Next, the data and research methodology are presented. The results are then presented, which is followed by the conclusion.

Literature Review

Recognition of the problem of real estate licensee turnover is not new. Over the two decades since Klimoski and Mitchell (1980) proposed a testing device to predict new real estate licensee success, numerous studies have examined factors associated with real estate licensee success. Several studies have examined service quality, including Johnson, Dotson and Dunlap (1988), Johnson, Nourse and Day (1988), Nelson and Nelson (1991), Coleman and Larsen (1995) and Okoruwa and Jud (1995). Other studies apply the human capital theory developed by Mincer (1974) to real estate brokerage. This theory holds that an individual's earnings are a function of a number of factors such as the individual's education and experience. Examples of these studies include Abelson, Kacmar and Jackofsky (1990) who examined licensees in Texas, Follain, Lutes and Meier (1987) who examined licensees in Illinois, Glower and Hendershott (1988) who studied licensees in Ohio, Sirmans and Swicegood (1997, 2000) who applied the model to samples of licensees in Florida and Texas and Crellin, Frew and Jud (1988) who applied the model to a national sample. A summary of these studies is presented in Exhibit 1. While a consensus does not exist, generally, these studies conclude that several of the characteristics that result in success in the brokerage business are similar to the characteristics that result in success in other fields of endeavor. For example, the number of hours worked, experience and the educational level attained by the agent tend to be positively related to success. None of these studies focus on new licensees.

Data and Research Methodology

The sample examined in this study consists of new licensees who joined either the Dayton Area Board of REALTORS (Dayton Board), the Stark County Association of REALTORS (Stark Association), or the Mansfield Board of REALTORS (Mansfield Board) between 1996 and 1998. The three areas selected provide geographical dispersion across the state and a cross section of board size.4 The Dayton Board serves the counties of Montgomery, Greene, Darke and Preble in southwestern Ohio. The Stark Association serves the counties of Stark and Carroll in northeastern Ohio. The Mansfield Board serves Richland County in central Ohio. General economic conditions were favorable and did not differ significantly between the three areas during the study period. Additional information about each area is provided in Exhibit 2.

Enlarge 200% Enlarge 400%

There are two basic steps in the methodology. The purpose of the first methodological step is threefold. The first is to provide some evidence of the extent of new licensee turnover. The second is to determine if new licensee survival is significantly related to three factors: licensee gender, employing firm size and franchise affiliation. The third is to determine if survival rates differ significantly between boards. Data for this portion of the study were gathered from several sources. The names of new licensees to the Dayton and Mansfield Boards (and their original broker affiliation) were obtained from the local boards. Because the Stark Association did not have a list of new licensees, the names of their new licensees (and each licensee's broker affiliation at the end of the calendar year) were obtained from the Ohio Association of REALTORS. Each list of new licensees was then compared to the membership directory of the local boards, published at the beginning of each of the following three years. For example, the membership directories for 1999, 2000 and 2001 were used to track a licensee that joined a board anytime during 1998. This procedure determined the status of broker affiliation of each licensee for the period. The membership directories were also used to determine the total number of licensees employed at each firm. For the purpose of this analysis, a firm was classified as small if it employed twenty-four or fewer licensees, and large if it employed twenty-five or more licensees.5 The local boards were contacted to ascertain whether each firm was a franchise affiliate or an independent operation and to confirm licensee gender.

An advantage of the process employed in this portion of the study is that it allows inclusion of all new licensees in the subject boards (data for the second step is limited to survey respondents). A disadvantage of this process is that, due to data limitations, only a few factors can be investigated. As the previous section makes clear, several studies have investigated the relationship of licensee gender, employing firm size, franchise affiliation (and other factors) with licensee income. While the focus of the present study is on licensee survival rather than licensee income, it is plausible that the two are positively correlated, and therefore, the findings can be compared with the findings of previous studies.

The data are used to calculate survival rates. First, survival is defined as the agent remaining in the local board. For the Dayton and Mansfield Boards, four different sets of survival rates are calculated: an overall rate, a rate based on the licensee's gender, a rate based on employing firm size and a rate based on whether the employing firm was a franchise affiliate. For the Stark Association and the entire sample (the three boards combined), only the overall rate and a rate based on licensee gender are calculated. Rates based on the other criteria could not be performed because the size and franchise affiliation of the original broker for nineteen new licensees could not be determined (recall the database for the Stark Association did not include each licensee's original broker). Apparently these licensees had exited the board before the end of the calendar year in which they entered the business. Survival rates are calculated at three different time intervals, which for expository expedience is referred to as a one-, two- and three-year rate.6 A variety of statistical tests are conducted. A chi-square test is employed to determine if there are significant differences in survival rates based on gender, and for the Dayton and Mansfield Boards to determine if there are significant differences in survival rates based on firm size and franchise affiliation. An ANOVA is performed to determine if survival rates differ significantly between boards.

Enlarge 200% Enlarge 400%

The above process is repeated with survival defined as the agent remaining with his or her original broker. This analysis may be of more interest to brokers who bear new licensee training costs. In this iteration, any new licensee that worked for a firm that ceased operation during the study period (either because the firm closed or merged with another) was eliminated from the sample because it would be impossible for a licensee to remain with the same (named) firm, and including the licensee would bias the survival rates downward.

Summary information about the number of new licensees and the firms that employed them is presented in Exhibit 3. Note, for example, that there were 1,106 new licensees in the three markets during the 1996-1998 time period, who were initially employed at one of 211 different firms. At some time between 1996 and 2001, 38 of the firms that initially employed a member of the sample ceased operations, and the 116 affected licensees were eliminated from the sample in the second iteration of survival rate calculations.

The second basic step in the methodology is to identify licensee characteristics that distinguish surviving licensees. To accomplish this task, a 41-question survey was mailed to 627 persons who passed the Ohio real estate salesperson's examination in 1996, 1997 or 1998, and joined the Dayton Board, the Stark Association or the Mansfield Board. The individuals to whom the survey was mailed were identified through records supplied by officials at the local boards. Three hundred ninety-six surveys were mailed to individuals that joined the Dayton Board, 56 to individuals who joined the Mansfield Board and 175 to individuals who joined the Stark Association. Two follow-up mailings were made in an attempt to increase the response rate; one direct to individuals who did not respond to the first request, and another to brokers to encourage licensees to respond to the survey.7

Once the survey data was obtained, tests were conducted to identify variables that distinguish licensees that survive for three years.8 For quantitative variables (e.g., licensee age), a two-sample t-test was performed to identify variables that differ significantly between licensees who remained in the local board and those that exited the board, as well as licensees that stayed with their initial broker and those that left the broker (either switching broker affiliation or exiting the local board before three years).9 An [alpha] = .05 was used to test if the unknown variances were equal. When the two samples had unequal variance, Cochran's approximation for the degrees of freedom was used to determine the critical t-value. When the two samples had equal variances, the pooled t-test for two populations with equal variances was employed. For qualitative variables (e.g., gender), the non-parametric chi-square test was employed to determine if any significant differences exist between the same two licensee groupings described in the preceding paragraph.10 Finally, a two-factor ANOVA was conducted on several quantitative variables to test for significant differences between boards.

Enlarge 200% Enlarge 400%

Results

Survival Rates-Licensee Remains in Local Board

Survival rates, with survival defined as the agent remaining in the local board, are shown in Exhibit 4. Information for the entire sample is presented in the first portion of the exhibit followed, in order, by information for the Dayton Board, the Stark Association and the Mansfield Board. The overall survival rates shown in the first portion of the exhibit provide evidence of the high turnover rates that concern industry leaders; only 76.7% of all new licensees in the sample were still active at the end of the calendar year in which they joined a board. The survival rates at the end of the second and third calendar years drop to 58.0% and 47.5%, respectively.

Enlarge 200% Enlarge 400%

Survival rates by gender are also presented in Exhibit 4 for the entire sample and for each of the three markets. Examination of the exhibit reveals that survival rates for female licensees are generally lower than for male licensees. But, the only difference that is statistically significant (at alpha = .10) is the one-year survival rate for the Dayton Board where 79.6% of male licensees survived till the end of the calendar year in which they joined the board compared to only 74.3% of new female licensees. The difference loses statistical significance by the end of the second calendar year and no significant difference in survival rates based on gender was detected for the Stark Association or Mansfield Board.

Previous research on the relationship of licensee gender and income provide contradictory results. Several studies discovered that male licensees earn significantly more than female licensees (Crellin, Frew and Jud, 1988; Glower and Hendershott, 1987; and Sirmans, and Swicegood, 1997, 2000). But, Abelson, Kacmar and Jackofsky (1990) found the opposite result. The findings of the present study, however, are more consistent with the findings of Follain, Lutes and Meier (1987) who did not discover a significant difference in earnings attributable to gender.

Survival rates based on the size of the firm that initially employs the new licensee are also presented in Exhibit 4 for the Dayton and Mansfield Boards. Statistically significant differences (at alpha = .10) were discovered in three cases. The percentage of new licensees in the Mansfield Board surviving until the end of the first calendar year was 83.7% for those employed at a small firm compared to 69.0% for those employed at a large firm. However, this difference disappears by the end of the second calendar year. In the Dayton Board, new licensees employed at a large firm had a significantly higher three-year survival rate; 49.7% compared to 41.7% for those employed at a small firm.

Survival rates for the Dayton Board are consistent with the finding of several previous studies that licensee income is positively related to firm size (e.g., Crellin, Frew and Jud, 1988; Follain, Lutes and Meier, 1987; and Sirmans and Swicegood, 199)). Survival rates in the Mansfield Board, however, are most consistent with the findings of Sirmans and Swicegood (2000) who did not find a significant difference in earnings attributable to firm size.

Survival rates based on whether the employing firm is associated with a national franchise are also presented in Exhibit 5 for the Dayton and Mansfield Boards. A significant difference is discovered in only one case. The percentage of new licensees in the Mansfield Board surviving until the end of the third calendar year was 46.2% for those employed at an independent firm compared to only 23.9% for those employed at a franchise affiliate. No significant difference in survival rates based on franchise/independent status is detected in the Dayton Board.

Previous studies find mixed results regarding the effect of franchise affiliation on licensee income. Sirmans and Swicegood (1997) find a positive effect on earnings associated with franchise affiliation. Survival rates in the Mansfield Board, however, are more consistent with the negative effect discovered by Crellin, Frew and Jud (1988), and survival rates in the Dayton Board are consistent with the findings Sirmans and Swicegood (2000) who detected no significant relationship between earnings and franchise affiliation.

Survival Rates-Licensee Remains with Original Broker

Survival rates, with survival defined as the licensee remaining with his or her original broker, are shown in Exhibit 5. Survival rates calculated under this definition are noticeably lower than under the previous definition. Compare, for example, the overall rates for the entire sample shown in the first portion of the exhibit with the comparable rates in Exhibit 3. Only 72.8% of the sample new licensees remained with their original broker until the end of the calendar year in which they joined a board compared to 76.7% who remained in the board after the same passage of time. The portion of new licensees that stayed with their original broker at the end of the second and third calendar years drops to 54.8% and 38.5%, respectively compared to 58.0% and 48.5%, respectively, of licensees that remained active in the board.

Enlarge 200% Enlarge 400%

Examination of Exhibit 5 reveals generally lower survival rates for female licensees than male licensees. But, only the one-year rate is statistically significant for the entire sample with a 77.2% male survival rate compared to 69.4% female survival rate. This difference appears to be driven by the significant difference in the one-year survival rate, based on gender, in the Dayton Board where 76.2% of male licensees remained with their original broker until the end of their first calendar year compared to only 65.3% of female licensees. Note that the difference loses significance in subsequent years and that no significant difference is discovered in the other two boards. Again, the results in our sample are consistent with the findings of Follain, Lutes and Meier (1987) who did not discover a significant difference in earnings attributable to gender.

Further examination of Exhibit 5 reveals only one statistically significant difference in survival rates based on the size of the firm that initially employs the new licensee. The percentage of new licensees in the Dayton Board that stayed with their original broker at the end of the third calendar year was 35.5% for those employed at a small firm compared to 28.7% for those employed at a large firm. No significant difference is detected in the Mansfield Board. This result is most consistent with the findings of Sirmans and Swicegood (2000) who did not find a significant difference in earnings attributable to firm size.

Survival rates based on whether the employing firm is associated with a national franchise are also presented in Exhibit 5. No statistically significant differences were discovered, which is consistent with the findings Sirmans and Swicegood (2000) who find no significant relationship between earnings and franchise affiliation.

Survival Rates-Differences between Boards

Cursory examination of the overall survival rates for the three boards in Exhibits 4 and 5 suggests that under either definition of survival new licensee survival rates in the Stark Association are higher than the other two boards. To formally test whether survival rates in the three areas are significantly different, a two-factor ANOVA was conducted on the overall survival rates in each board, and the least significant difference (LSD) was used for Post Hoc analysis. The results, summarized in Exhibit 6, indicate that under either definition of survival, survival rates for new licensees during the study period in the Stark Association were significantly higher than the other two boards. No statistically significant difference between the Mansfield Board and Dayton Board (at alpha = .05) is indicated.11

There are three possible explanations for the significantly higher survival rates for new licensees in the Stark Association. One possibility is a sampling problem (Type I error), which is beyond the control of the researchers. Another possibility is a nonsampling error, but the process was checked and rechecked to ensure the accuracy of the analysis. While the possibility that there are errors in the data cannot be discounted, the people who supplied the data took the task seriously. A third possibility is that the model is correct and there are logical reasons that explain the difference in survival rates.

To test the third possibility, the following information was obtained for the years 1996 through 2001. Each board supplied annual information on the total number of licensees at year-end (both sales associates and brokers), the total number of residential properties sold and the total selling price of those units. In addition, annual population and the number of households in the market area served by each board were obtained from various issues of Demographics USA County Edition (1997-2002). Using this information, five measures were calculated for each year for each board: residential units sold per number of households, residential units sold per licensee, sales dollars per licensee, population per licensee and number of households per licensee. The higher (lower) any of these measures, the easier (more difficult) it would be for a new licensee to survive, ceteris paribus. It is unfortunate that full-year data was not available for the Stark Association for each year. But partial-year data submitted by the Stark Association for the years in which full year data was unavailable suggests that the missing annual figures do not differ dramatically from years where full-year data was available.

Enlarge 200% Enlarge 400%

Exhibit 7 contains the information and measures described above. The Stark Association is the only board in the sample with increased membership between 1996 and 2001. Over that time period, the membership in the Stark Association increased by nearly 61% while the Dayton and Mansfield Board memberships decreased by 9% and 8%, respectively.

A one-factor ANOVA (with Post Hoc analysis utilizing the LSD test) was conducted on each of the five measures described above. The results are summarized in Exhibit 8. Note that no statistically significant difference is discovered for the percentage of units sold per household between the Stark and Mansfield markets (i.e., property turnover is roughly the same in each of these markets). But, a significantly larger percentage of properties were sold per household in the Dayton market compared to the other two. In essence, members of the Stark Association have no comparative advantage with respect to this measure. In fact, they (as well as members of the Mansfield Board) are at a significant disadvantage compared to Dayton. Despite this disadvantage, every other measure is significantly higher in the area served by the Stark Association compared to the other two boards. This suggests that it is easier for a new licensee in the Stark Association to survive compared to a licensee in the other boards, and this may account for the higher survival rates for new licensees observed in that region.

Enlarge 200% Enlarge 400%

Licensee Survey

Summary survey response information is provided in Exhibits 9 and 10. Response rates by board and licensee group are shown in Exhibit 9.12 Note that the overall response rate was 23.4%. Exhibit 9 contains response information about selected quantitative variables. Exhibit 10 contains response information for qualitative variables. In Exhibit 9, the number of respondents that reported each variable for the total sample and for each board, along with the average value of each variable, are reported. In Exhibit 10, the figures shown represent the number of respondents reporting each variable and the percentage of respondents reporting the variable.

The test results for quantitative variables are presented in Exhibits 11 and 12. In Exhibit 11, licensees that remained in the local board are compared to those that did not. Note (in the last column) that most variables are not significant at the 5% confidence level, but two variables-previous sales experience and number of days till first listing-were nearly so. In Exhibit 12, licensees that stayed with their original broker are compared to those that did not. Again, most variables that were investigated are not significant, but the number of years in previous occupation was nearly significant.

Regardless of the manner in which licensees are grouped, three variables are statistically significant: previous annual income, number of dependents and first year income. Two of these could be used as a screening device because they are observable before the prospective licensee is sponsored: previous annual income and number of dependents. Specifically, the mean income earned during the year preceding entry into the real estate business is significantly higher for licensees who remained in the board compared to licensees that exit, and it is also higher for licensees who stay with their original broker compared to those that left. It appears that people with a successful work history and responsibility for the financial welfare of others have a higher probability of survival in the board, and with their original broker. First year income, and in Exhibit 12 only, first year average hours worked cannot be used as a screening device. The survey results with regard to these variables are generally in line with intuition (e.g., first year income for licensees that stayed with their original broker is significantly higher compared to those that did not).

Enlarge 200% Enlarge 400%

Enlarge 200% Enlarge 400%

Enlarge 200% Enlarge 400%

Enlarge 200% Enlarge 400%

The results of the chi-square test on qualitative variables are presented in Exhibits 13 and 14. For expository expedience, statistically significant variables are presented in Exhibit 13 and insignificant variables in Exhibit 14 (shown in the Appendix). In these exhibits, the test results comparing licensees that remained and exited the board are presented in the third and fourth columns, and the test results comparing licensees that remained with and left their original broker are presented in the fifth and sixth columns. The significant variables are discussed below.13

Marital Status. Eighty-nine respondents reported that they were married at the time they entered the real estate business. Analysis of the survey results indicates that individuals are more likely to remain in the board and with their original broker if they are married. One hundred percent of married licensees remained in the board compared to 56% of unmarried individuals. Sixty-six percent of the married respondents remained with their original broker compared to 48% of unmarried individuals.14

Learned of Position through Friends or Relatives. Fifty-three respondents reported that they learned about the position in real estate through a family member or friend. Seventy-seven percent of these remained in the board compared to 100% of those that did not indicate this response. Only 52% remained with the original broker for three years, while 80% of those that did not learn of the position in this manner remained with their original broker.

White-Collar Job. Seventeen respondents indicated that one motivation for entering the business was that they wanted a white-collar job. Forty-one percent of these individuals remained in the board compared to 88% of the respondents who did indicate this motivation. Only 29% of these individuals stayed with their original broker compared to 68% of those that did not specify this motivation.

No Opportunity for Advancement in Previous Job. Nineteen respondents reported that a motivation for entering the real estate profession was a lack of advancement opportunities in their previous employment. Sixty-five percent of them remained in the board compared to 87% of those that did not specify this motivation. This variable was not statistically reliable for comparing licensees that stayed with or left their original broker.

Bachelors Degree. Forty-one respondents reported that they held a Bachelors degree at the time they entered the profession. This variable was insignificant between those that remained and left the board. However, 75% of these individuals remained with their original broker for three years compared to 41% that did not hold a degree.

Office Close to Home. Thirty-eight respondents indicated that a consideration in selecting a broker was that they wanted an office located close to their home. Seventy-one percent of these individuals remained in the board compared to 87% of those that did not indicate this motivation. Only 51% of the individuals that specified this motivation remained with their original broker compared to 67% that did not specify this motivation. This indicates that when convenience is a motivational factor for the licensee, chances of survival diminish.

Enlarge 200% Enlarge 400%

Change in Marital Status. Forty-two respondents reported a change in marital status during their first three years in the profession. Forty-eight percent of the respondents who indicated this event remained in the board compared to 98% of those whose marital status did not change. Individuals in the sample who experienced such an event were nearly twice as likely to leave the broker; 66% left.

Five of the significant variables identified involved differences between new licensee employment expectations and realizations: independence, flexible hours, income, business expenses and advancement opportunities. Careful reading of the following paragraphs will make it obvious that licensees with realistic expectations are more likely to remain active licensees and with their original broker.

Independence. Twenty-six respondents reported that the degree of independence that they were afforded in their position in real estate was less than expected. Forty-three reported that their position afforded them more independence than expected, and 78 reported that their expectations were met. Only 50% of those whose realizations were less than expected remained in the board compared to 88% of those whose realizations were higher than expected, and 90% of those whose expectations were met. Thirty-one percent of those whose realizations were worse than expected remained with their original broker compared to 63% whose realizations were better than expected, and 74% whose expectations were met. To decrease turnover, if a broker discovers that the prospective employee has the desire to work independently, the broker should take steps to insure that this happens, or (if the broker has a more regimented system) not sponsor such individuals.

Flexible Hours. Thirty-four respondents reported that their working hours were less flexible than anticipated. Thirty-nine reported more flexible working hours than anticipated, and 74 reported that their expectations were met. Sixty-five percent of those who realized less flexible hours than anticipated remained in the board compared to 92% who realized more flexible hours than expected and 85% of those expectations were met. Thirty-five percent of those licensees in the first group stayed with their original broker compared to 64% in the second group, and 74% whose expectations were met. To reduce turnover, therefore, brokers should take steps to insure that prospective sales associates have a clear understanding of what to expect in this area.

Income. Forty-six respondents reported that their income was lower than expected. Thirty-eight reported that it was higher than expected, and 63 reported that their expectations were realized. Fifty percent of those that realized lower income than expected remained in the board compared to 97% in both the other groups. Thirty-three percent of the individuals in the first group stayed with their original broker compared to 68% of those whose income expectations were exceeded, and 82% for those whose expectations were met.

Business Expenses. Ninety-two respondents expressed their opinion that business expenses were higher than anticipated. None expressed the opinion that expenses were lower than expected, and 55 reported that their expectations were realized. Seventy-four percent of those who realized higher expenses than anticipated remained in the board compared to 96% of those whose expectations were met. Fifty-four percent of the first group stayed with their original broker compared to 78% of those who indicated expenses were in line with expectations. Brokers may be able to reduce turnover by clearly explaining to prospective sales associates the expenses that new licensees will incur.

Advancement Opportunity. Eleven respondents reported realizing less advancement opportunities than expected. Sixteen reported better advancement opportunities than expected, and 120 reported that their expectations were realized. Fifty-five percent of those who reported less opportunities than expected remained in the board compared to 94% of those who realized more opportunities than expected, and 83% whose expectations were realized. The results for this variable were not statistically reliable for comparing licensees that stayed with or left their original broker.

In interviewing prospective licensees, brokers generally try to elicit information that will suggest whether the prospect will be a good fit for the broker's company. The results indicate that several factors that would normally be considered "desirable" in a new licensee actually were not in the sample. Consider the following three variables.

Work with People. Sixty-eight respondents reported that a motivation of entering the real estate business was that they wanted a job where they could work with people. Seventy-five percent of the respondents who indicated this motivation remained in the board compared to 90% of those that did not express this motivation. Only 54% of the respondents who stated this motivation remained with their original broker for three years, compared to 71% who did not specify this motivation.

Challenging Job. Fifty respondents reported that a motivation for entering the profession was that they wanted a challenging job. Seventy-four percent of those expressing this motivation remained in the board compared to 88% that did not. Only 47% of respondents that specified this motivation stayed with their original broker for three years compared to 72% that did not specify this motivation.

Sales Job. Twenty-nine respondents indicated that a motivation for entering the business was that they wanted a position in sales. Sixty-nine percent of the respondents who indicated this motivation remained in the board compared to 86% of those who did not express this motivation. Only 46% of licensees that specified this motivation stayed with their original broker compared to 67% that did not specify this motivation.

Conclusion

The high turnover rate of newly licensed real estate agents is a concern to industry leaders, individual brokers and new licensees. Despite this concern, little formal research on this issue has been published. To help remedy this shortcoming, two methodologies are employed to investigate the issue. The first methodology involves the calculation of survival rates. For this portion of the study, the sample consists of all licensees new to the Dayton Board of REALTORS, the Stark County Association of REALTORS and the Mansfield Board of REALTORS during the years 1996 through 1998. An advantage of this methodology is that it allows inclusion of all new licensees in the subject boards. A disadvantage is that, due to data limitations, only a few variables can be investigated.

Survival was defined in two ways; first, as the licensee remaining in the local board/association. Under this definition, 76.7% of all new licensees survived to the end of the calendar year in which they joined. By the end of the second and third calendar years, the survival rate dropped to 58% and 47.5%, respectively. Second, survival is defined as the licensee remaining with his or her original broker. Under this definition, 72.8% survived to the end of the first calendar year, and by the end of the second and third calendar years the survival rate dropped to 54.8% and 38.5%, respectively. In addition to overall survival rates, survival rates based on three criteria were calculated and tested to identify significant differences. No persistent significant difference in survival rates based on licensee gender, employing firm size or employing firm franchise affiliation was discovered in the sample.

An unanticipated finding was that overall survival rates in the Stark Association were significantly higher than the other two boards. The difference appears to be due, at least in part, to the fact that there are significantly fewer total licensees per capita in the Stark Association market area. If this is the case, high turnover rates for new licensees may be symptomatic of a different problem. Perhaps there are more licensees in the Dayton and Mansfield Boards than needed to efficiently serve the market (and possibly in the Stark Association as well). A broader survival rate study of all real estate boards in Ohio could provide valuable additional evidence on this possibility.

There are several actions that the industry or individual brokers could do to enhance new licensee survival. A few brokers in the sample markets are abandoning the traditional independent contractor relationship with some promising new licensees and instead are hiring them as employees; transitioning them to a fully commission-based compensation system as they build their client base. Despite the fact that in the study the orientation and training variables were not significantly related to survival, effective formal training should also facilitate a licensee's transition to a commission-based compensation system. This could include compensated pre- or post-licensing apprenticeship programs. Some firms employ a noncompensated mentoring program, where an experienced agent supervises the activities of a new licensee. At least one broker in the Stark Association pays the mentor 10% of any commissions generated by the licensee being mentored. The success or failure of the various types of training/mentoring programs needs to be further investigated.

Another way to reduce the problem of new licensee turnover would be to increase entrance requirements. With fewer new entries into the industry, the natural attrition of existing licensees will reduce the number of total licensees to a level that increases the probability of survival for the new entrants. Increased entrance requirements could take many forms including requiring more prelicensing education (e.g., time management, salesmanship and business management courses) or specifying a higher minimum passing score on the sales associate examination. The increased competence of individuals that meet or exceed higher entrance standards should also enhance survival rates.

The second methodology employed involves the analysis of survey information obtained from 147 licensees new to the Dayton Board, the Stark Association or the Mansfield Board during the years 1996 through 1998. Tests were conducted to identify variables that differed significantly between licensees that remained active in the local board for at least three years and those that exited the local board, as well as licensees that stayed with their original broker for at least three years and those that did not. An advantage of this methodology is that it is possible to test a wide variety of variables. A disadvantage of this methodology is the possibility of response bias. In addition, contacting former licensees to participate proved problematic. If individual brokers, and or real estate authorities wish to collect data on the experience of licensees that exit the industry, it is suggested that a formal exit interview be conducted at the time the licensee switches to inactive status.

At least eight variables identified in this study could be used as an effective screening device. Several of these variables suggest that educated people with a successful work history and responsibility for the financial welfare of others have a higher probability of remaining in the board, and are also more likely to remain with their original broker. This was the case for licensees who held a Bachelors degree, were married, with relatively more dependents, with relatively high annual income in the year prior to entering real estate brokerage or with relatively more years of work experience. On the other hand, individuals who reported that a factor they considered in selecting a sponsor was an office location close to their home, that they learned of the position through a family member or friend, that opportunities for advancement were limited in their previous employment or that they wanted a white-collar job were more likely to leave the broker.

In interviewing prospective licensees, brokers generally try to elicit information that will indicate whether the prospect will be a good fit for the broker's company. The results suggest that several factors that would normally be considered "desirable" in a new licensee actually were not. Licensees that reported that (at least) part of their motivation for entering the real estate brokerage profession was to work: with people, in a challenging job or in a sales position switched broker affiliation or left the board at higher rates than those who did not specify these motivations.

The results also indicate that brokers can reduce new licensee turnover by ensuring that new licensees have accurate expectations regarding several variables. Specifically, it was determined that licensees who discovered that their duties entailed less flexible hours, less independence of action, more business expenses, less opportunities for advancement or less income than anticipated switched broker affiliation or exited the board at higher rates than other licensees.

Enlarge 200% Enlarge 400%

Enlarge 200% Enlarge 400%