According to information published by the National Association of Home Builders (NAHB) and based on the 1997 U.S. Census of Construction Industries, most of the general residential contractors hiring employees at that time -- some 79,100 companies -- were mainly engaged in producing new homes. Of these new home builders, more than 73,500 firms (93 percent) were small-volume home builders producing fewer than 25 homes per year. These firms accounted for 38.8 percent of all homes started during the year (Willenbrock, 2001). By the spring of 2000, at the time this study was proposed, the number of residential builders with employees that concentrated mainly on new home production had increased to approximately 90,000 (Carliner, 1999). Once again, the majority of these builders were small-volume home builders. At the time this study was conducted, these home builders could expect to generate average gross profit margins of approximately 14 to 26 percent and average net profit margins before income taxes of approximately 3.5 to 5.5 percent (Maltzman, 1998; NAHB, 2001; RMA, 2000; Shinn, 1995).

In the most recent comprehensive financial survey published by the NAHB, the average small-volume home building company in the lower quartile (bottom 25 percent of those surveyed) barely made enough profit to break even, showing an average net profit margin of 1.1 percent. The average gross profit margin for these lower-quartile companies was 13.2 percent. Similar-sized companies in the upper quartile (top 25 percent) enjoyed net profit margins of 10.8 percent, while the average gross profit margin for those in the upper quartile was 21.2 percent (NAHB, 2001).

Based on the differences in profit margins between companies in the upper-quartile and those in the lower quartile, it stands to reason that something different must be happening between the two groups. To date, much research has been published regarding how to identify the current health and profitability of businesses, including home building companies, by analyzing their financial ratios and comparing them to established industry standards. However, another approach to assessing the vitality and health of a business would be to determine industry standards with regards to management practices and to examine these non-financial indicators as possible predictors of profitability.

Owners of small businesses, including general building contractors, that experience failure often try to place the blame on external influences, but this argument has been rejected by several studies which report that some small businesses, including home builders, can make record profits even when times are tough. In fact, it has been shown in some cases that superior management skills and abilities often allow owners to navigate their way to success, even in bad times (Evans, 1992; Flahvin, 1985). Because of the many administrative activities and management requirements placed upon home builders, one of the most serious obstacles to success is seen as the implementation of questionable management systems and controls in a company, or maybe worse, the complete lack of any established control systems within a company (Flahvin, 1985; Gaskill, Gill, 1968; Strischek, 1998). All home building companies need management systems and procedures, but small-volume home building companies, with limited administrative staffs, are probably the most prone to failure without sound, proven systems and procedures (NAHB, 1979).

By designing a scientific study using a representative sample of small-volume home builders nationwide, it should be possible to measure the frequency of use of specific management practices and to identify those practices that are common to highly profitable companies. In addition, the study should be designed to differentiate between management practices that characterize successful companies and those that characterize unsuccessful companies.

The purpose of this research was to determine which, if any, of the pre-selected management practices identified in small-volume home building companies were significant indicators of the current level of profitability of those companies.

The first subproblem was to generate a list of management practices which were considered by industry experts to be the most important ones for small-volume home builders. The second subproblem was to develop a survey instrument that would effectively measure the level of use of each of the selected management practices in the companies participating in this study. The third subproblem was to determine the profit margins of each of the companies surveyed. The fourth subproblem was to determine the relationship between companies’ management practices and their profit margins and to identify which practices were highly correlated to the profitability of small-volume home building companies.

The main purpose of this study was to determine which of the management practices identified were significant predictors of profitability for small-volume home builders. The hypothesis for this study follows:

H_o: There are no differences in the management practices used by successful and unsuccessful

small-volume home building companies.

H_a: There are differences in the management practices used by successful and unsuccessful

small-volume home building companies.

This research was intended to explain some of the most important management practices which contributed to the success of small-volume home building companies; however, it was not intended to explain all possible factors which might influence their success. Because there are many factors that might affect the well-being of a home building company, it would be impossible to account for all of them in one study. For example, such things as leadership abilities of owners, inherited wealth of owners and their reputation in the community are but a few of the possible confounding factors that might contribute to the success or failure of a company.

This study was intended to identify the profit margins of responding companies as dependent variables for the statistical tests described within to identify which companies were successful and which companies were unsuccessful. This study was also intended to identify levels of use of management practices and the correlation of some specific management practices on profit margins only in the identified population of companies, which are members of the NAHB. It is possible that the same principles that apply to these NAHB companies also apply to other small-volume home builders which are not members of the NAHB, but this cannot be concluded from the results of this study.

To begin this study, a broad list of more than one hundred management practices was developed using a three-fold approach. First, the professional background of one of the authors, including more than twenty years as an owner and operator of small-volume home building businesses in addition to teaching construction company management classes at the university level provided a bird’s-eye view of the problem; second, literature related to management practices adopted by home builders was carefully reviewed; and third, owners of home building companies, certified public accountants dealing with clients in the home building industry, and university faculty members whose areas of research address issues in construction company management were interviewed. After careful review by a committee of some fifteen members of the NAHB’s Business Management and Information Technology Committee, the list was narrowed to approximately fifty management practices that were considered to be most important to small-volume home builders, and a questionnaire was developed to measure the level of use of each practice by companies to be surveyed. During a period of several months, the questionnaire was carefully reviewed by university faculty members of construction management programs, construction management students, and most importantly, by home builders to test for understanding and readability. Throughout the review process, appropriate revisions were made until the final questionnaire was approved by the NAHB’s National Business Management and Information Technology Committee.

In this study, the population of interest consisted of small-volume home builders in the United States, members of the NAHB, where new home sales represented the majority of each company’s business. According to the most recent information published and available immediately prior to the date of the study, there were approximately 62,450 builder members of the NAHB nationwide, including home builders, remodelers, and developers. Of these, some 40,984 firms reported that they started at least one new home during the year. Companies that reported starting 1 to 25 new residential units for the year numbered 27,542 and represented more than two-thirds of the firms that reported housing starts (Bajwa, 2001).

As noted above, the population of interest for this research was quite large. In an attempt to obtain results that would provide a 95 percent level of confidence in a national study, with a plus or minus five percent margin of error, this study was designed to obtain data from at least 400 firms (Weisberg & Bowen, 1977). A systematic random sample of 1,112 companies was obtained from the NAHB’s data base of member builders reportedly building 11 to 25 new homes per year. This was accomplished by first stratifying by zip codes a list of approximately 21,120 companies provided by the NAHB. As the questionnaires were received, it was discovered that the majority of those companies actually built fewer than 10 homes per year. Because of this, the results of this research focus on the data received, which includes companies that built 1 to 25 new homes per year. To reach the sample size indicated above, every 19^th company from the stratified list was selected as a candidate. The sampling procedures used in this study were primarily patterned after Don A. Dillman’s Total Design Method and consisted of reminder post cards and additional mailings of the complete survey instrument to non-respondents in order to maximize response rates (Dillman, 1978). By following these steps, 478 questionnaires were returned prior to processing the data, representing a response rate of approximately 43 percent of the total sample. This was approximately three to four times higher than average returns for management studies conducted during the last few years by the NAHB’s Business Management Committee (Freedman, 2000).

For purposes of applying the statistical tests discussed in this report, only questionnaires from companies that built 25 and fewer new homes per year where revenues from new homes represented more than 50 percent of their revenues were used. Based on these procedures, of the 478 questionnaires that were returned, 359 were considered responsive for companies building 1 to 25 new homes per year.

In order to try to correlate the independent variables -- represented by the management practices addressed in the questionnaire -- to the dependent variable of profitability, it was first necessary to determine which companies actually provided profit figures that were usable. The study was originally structured to derive both gross profit margins and net profit margins, but because few companies provided reliable information regarding their net profit margins, a decision was made to use gross profit margins as the only dependent variable upon which inferential statistical tests were performed. According to NAHB staff personnel and certified public accountants serving the NAHB in various capacities, determining net profit margins for small-volume home builders has also been difficult in previous studies (Freedman, 2000; Shinn, 2000; Maltzman, 2000).

Not all of the 359 qualified companies building 1 to 25 new homes provided enough financial information to derive reliable gross profit margins. Of the 359 companies, only 236 companies provided data necessary to perform the needed gross-profit-margin calculations. Because gross profit margins provide funds necessary to pay the owners’ salaries and all general and administrative overhead costs for a company, it was considered unreasonable to include the 23 companies reporting negative gross profit margins for purposes of this study. Zero gross profit margin was chosen as a cutoff point because those companies were literally trading income dollars for hard costs of construction, a sure recipe for operational insolvency. Following this decision, a dot plot was created that described the remaining 213 companies (see Figure 1). Notice that there were several possible "breaks" in the data points that could arguably be used to eliminate companies with very high gross-profit-margins. According to financial studies targeting small-volume home builders published by the NAHB and other groups interested in these ratios, gross profit margins typically average somewhere between 15 and 30 percent. It is obvious that those data points exceeding 90 percent are unreasonable outliers, probably resulting from inaccuracies in the financial reporting. It is also obvious there are other breaks in the data occurring at about 38 percent, 42 percent and again at about 48 percent. Arguments could be made that breaking the data at any one of these points would be reasonable. However, for purposes of this study, a decision was made that any gross-profit margin observations exceeding 50 percent were to be eliminated from the sample. This not only represents a major break in the data points as seen in the dot plot, but it can also be argued that this is a reasonable choice from the standpoint that some small-volume home building companies might show fairly high gross profit margins because of several accounting factors, including owners who work on the jobs without taking wages as hard costs. Based on this decision, the 13 companies with gross profit margins greater than 50 percent were also eliminated from the sample.

After applying the filtering requirements for gross profit margins discussed above, the final sample size for inferential statistical tests consisted of exactly 200 companies. The original goal of this research was to obtain at least 400 usable responses in order to limit the margin of error for testing purposes to a maximum of 5 percent. Although more than 475 questionnaires were actually returned, because of the filtering methods explained, the final number of usable questionnaires was 200, resulting in a maximum margin of error for testing purposes of 7.2 percent (Weisberg & Bowen, 1977).

Of the 1,112 companies in the original sample, 478 returned questionnaires. Telephone calls were made to the owners of every tenth company not responding to the survey. Based on answers received to predetermined sorting questions, it was evident that the companies contacted were in fact similar to those companies that had responded to the survey. Although the follow-up to the non-responses was not necessary, the results described above lead the researchers to believe that had the non-responding companies filled out and returned questionnaires, their answers would have been similar to those companies that in fact responded to the survey.

The data collected from this study were analyzed using two different statistical methods, correlation coefficient (Pearson’s) tests and contingency table analysis. The correlation tests were intended to find any variable that was significantly related to gross-profit margins, while the contingency table analysis was used to find the individual management practices that related to highly profitable versus less profitable companies. All data were analyzed by using SAS statistical software, Version 8.

By looking at Figure 2 below, a histogram showing the sample observations for gross profit margins for companies building 1 to 25 new homes per year, it is obvious that the dependent variable, gross profit margin, was not normally distributed. Because of the non-normality of the data, a non-parametric approach was selected for the correlation analysis performed using rank-ordered gross profit margin as the dependent variable in all cases.

Pearson’s correlation is a statistic that looks for relationships between variables. In the context of this study, rank-ordered gross profit margin is the dependent variable. Each of the 101 independent variables was compared to the dependent variable in order to determine whether there was a significant correlation between any one of the independent variables and the dependent variable. Because so many factors were tested, it was necessary to divide the desired overall alpha value (p < .1) by the number of tests performed on all variables (101) to determine which variables were truly significant. The resulting level of significance for testing purposes of individual variables was p < .001.

The research hypothesis for the Pearson’s correlation statistic can be stated as follows:

H_o: There are no relationships between rank-ordered gross profit margin and each of the

variables tested.

H_a: There is a relationship between rank-ordered gross profit margin and each of the variables

tested.

The second test that was performed was contingency table analysis (the chi-square test for independence). For purposes of this test, successful companies were defined as all those that were within the upper quartile, based on rank-ordered gross profit margin. Unsuccessful companies were those that fell into the lower quartile, also based on rank-ordered gross profit margin. The chi-square test looks at individual variables to determine if there is a difference between how profitable and unprofitable companies responded to each individual question posed in the survey.

Once again, an overall alpha level of .1 was used to determine the significance of the chi-square tests. However, just as with the single-variable correlations described above, it was necessary to divide the overall alpha level of .1 by 101 (the number of tests in this case) to arrive at the true alpha level, essentially p < .001.

The research hypothesis for this portion of the study follows:

H_o: The response to each individual question on the survey is independent of whether a

company is successful or unsuccessful (by definition).

H_a: The response to each individual question on the survey is related to whether a company is

successful or unsuccessful.

This section discusses the findings from the inferential statistical tests discussed in the section on Methodology. The findings provide information regarding relationships between the management practices tested and the profitability of small-volume home building companies.

A Pearson’s Correlation Coefficient test was run on 101 independent variables with the dependent variable, at a true alpha level of .001, to the performance of the continuous dependent variable, rank-ordered gross profit margin. Three variables seemed to have a definite relationship to profitability for companies starting 1 to 25 new homes per year. They are shown below in Table 1. Fifteen other variables had p-values < .1, but because of the number of tests that were run, it cannot be stated with certainty that they have a significant relationship to profitability. The following table lists the results of the single-variable correlations for companies starting 1 to 25 new homes annually. All questions that had p-values < .1 are included, along with a description of the question, the actual p-value, and the Pearson’s correlation coefficient. The variables are listed in order according to p-value, beginning with the most significant. Note the positive or negative correlation for each variable described.

Based on Pearson’s correlations, three variables are significantly related to profitability within companies. The correlations are not one-to-one, but they give us directionality. From this we conclude that these variables tend to have an effect on profitability, either negatively or positively. If the practices that relate positively to profitability are employed, will the company be more profitable? Not necessarily, but based on the results of this test, a company would tend to be more profitable if it does them.

In conclusion, it is impossible to state that the independent variables listed above, with p-values > .001 are statistically significant. From a statistical point of view, this is not surprising, given the number of tests performed. It appears that profitability is most likely a function of many variables. Although all the management practices listed above were correlated in some way to profitability, paying higher sales commissions and building homes on speculation were the only statistically significant practices that related positively to profitability. The only practice that significantly related to profitability in a negative direction was building new homes on building lots provided by customers.

The chi-squared test for independence in this study was applied to determine whether a correlation existed between the gross profit margin of each company and each independent variable, taken one at a time. If variable A, any one management practice, is not related to the quartile to which the company belongs, there should be no relationship. The sample of companies building 1 to 25 new homes per year included 200 questionnaires. For purposes of this report, the companies were grouped by quartile. The top 50 companies represented the upper quartile, and the bottom 50 companies represented the lower quartile. The middle 100 companies, representing the second and third quartiles, were not considered in this test. Of those questions that tested for significance at an overall level of p < .1, only one of them was truly significant at p < .001. As previously explained, this is because the alpha level was once again affected by the number of tests (101) that were performed. Only those variables that tested for significance at a level of p < .1 are discussed in Table 2. The variables listed may indicate a relationship to the profitability quartile as mentioned earlier, but except for companies that built a good share of their homes on speculation, that relationship cannot be considered statistically significant. Note that the variables are listed in order below according to p-value, beginning with the most significant.

Although all the management practices discussed above were correlated in some way to companies in the two quartiles, the only statistically significant practice reported from the chi-square test for independence was the fact that companies in the upper quartile sold more new homes on speculation than did those in the lower quartile.

Based on the inferential statistical tests described in this study, only three management practices were significantly correlated to profitability. There was a positive relationship between profitability and companies that built a good percentage of homes on speculation. There was also a positive relationship between profitability and companies that paid high sales commissions to market their new homes, indicating significant broker involvement at point of sale. The only statistically significant practice that was negatively correlated to profitability was when companies built new homes on customers’ lots. These findings all seem to indicate that successful companies were not building as many presold homes for specific owners as were the less successful companies. Rather, the successful companies were marketing a good portion of their homes to owners after construction was already begun.

It should be noted that this has been an investigative study into an important, but not thoroughly researched, topic in the home building industry. It appears from the findings that the problem of determining profitability in small-volume home building companies based on non-financial predictors is very dynamic. To say for sure that one or even a combination of management practices is solely responsible for profitability does not seem possible, given the evidence of this study. However, it can definitely be said that new information has been discovered and reported regarding the management practices of small-volume home builders in the United States. Another important note is, that according to the information provided by the respondents, this research was performed when the economy was generally very strong. It is possible that the results may differ during a less robust economy.

Bajwa, B. (2001, May 16, 2001). NAHB Data Processing Department.

Carliner, M. (1999, July 28, 1999). [Staff vice-president of NAHB Economics Department] ( Mark Hutchings, Ed.). Washington, D. C. (by phone).

Dillman, D. A. (1978). Mail and telephone surveys. New York, New York: John Wiley & Sons.

Evans, L. S. (1992). Readings in Management of the Home Building Business ( Lee S. Evans, Ed.) (p. 195 pages). Nederland, Colorado: Lee S. Evans and Associates, Inc.

Flahvin, A. (1985, October). Why small businesses fail. The Australian Accountant, pp. 17-20.

Freedman, A. (2000, 11 May). (Business committee meeting discussing questionnaire and profit margins). Washington, D.C.

Gaskill, L. R., Van Auken, H. E., & Manning, R. A. (1993, October). A factor analytic study of the perceived causes of small business failure. Journal of Small Business Management, pp. 18-31.

Gill, P. G. (1968). Systems management techniques for builders and contractors. New York, NY: McGraw-Hill, Inc.

Maltzman, S., C.P.A., Hays, S. W., C.P.A., & Freedman, A. (1998). 1997 cost of doing business study. Washington, D.C.: National Association of Home Builders.

Maltzman, S., C.P.A. (2000, 4 February). . Telephone conversation 909-798-8930.

National Association of Home Builders. (1979). Management manual for the small-volume home builder. Washington, D.C.: Author.

National Association of Home Builders. (2001). The business of building. Washington, D.C.: NAHB Business Management Department.

RMA. (2000). RMA annual statement studies 2000-2001. Philadelphia, PA: Author.

Shinn, C. C., Jr. (1995, May, June, July). Where did my profits go? How do I get them back? Part I. The Builder's Management Journal, 9(1), 1-4.

Shinn, C. C., Jr. (1995, August, September, October). Where did my profits go? How do I get them back? Part II. The Builder's Management Journal, 9(2), 1-4.

Shinn, C. C., Jr. (1995, November, December). Where did my profits go? How do I get them back? Part III. The Builder's Management Journal, 9(3), 1-3.

Shinn, E. (2000, 24 March). (Obtaining viable numbers for net and gross profit margins for homebuilders). Nashville, TN.

Strischek, D. (1998, July). Red warning flags of contractor failure. Journal of Lending & Credit Risk Management, 80(11), 40-47.

Weisberg, H. F., & Bowen, B. D. (1977). An introduction to survey research and data analysis. San Francisco, CA: W. H. Freeman and Company.

Willenbrock, J. H., P.E. (2001). Management guidelines for growth-oriented homebuilders. Arizona State University: Del E. Webb School of Construction (352 pages).