Chapter Listing | Return to 1997 Topical Index


Study #1, Organizational Level Research

The sample consists of firms that conducted their IPO in 1993. Although a total of about 700 firms are in the overall sample, the final sample for this study is 383 due to firms being dropped because they do not manufacture a good or service, they do not have any employees, or as a result of missing data.

As in the earlier research (Welbourne & Andrews, 1996), the prospectus of each firm was coded to obtain measures of employee value and several control variables. In addition, measures of firm performance were acquired from COMPUSTAT. An additional component of this study is the inclusion of a second measure of employee value obtained with a survey.

Employee Value

I use two different measures of employee value for this study. The first is the same measure used in the Welbourne and Andrews (1996) study. This variable attempts to assess the degree to which a firm values its employees and views employees as a strategic advantage. Because the prospectus describes a company's competitive advantages, it is a good source of information on the importance of employees as one of those potential advantages. The presence of a variety of indicators are coded and then summed. Those indicators of employee value are:

1. Whether the company's strategy statement and/or mission statement includes a reference to employees being a strategic competitive advantage.

2. The existence of a training program for employees.

3. The inclusion of an officer responsible for the human resource management function.

4. If the company used full-time employees, rather than part-time or temporary workers.

5. The SEC required rating of employee relations environment.

An additional measure of employee value was used for this study, and as you will see, the analyses are run separately for each measure.

A survey was sent to all members of the top management team listed in the prospectus of each firm. Thus, a total of approximately 5,000 surveys were sent to to p management at all the original 700 organizations. We received survey responses from 500 individuals, representing 356 organizations. The survey listed 40 different factors that were obtained from the literature on survival of IPO firms. The executive s were asked the degree to which each of these factors was important for their firm's performance since the IPO. The items dealt with the firm's overall strategy, financing, venture capital backing, product, technology, management team, IPO timing, and employee relations. The questions were based on the items that we used in the telephone survey to survivors in the 1988 IPO study (Welbourne & Andrews, 1996). The factor analysis resulted in a 5-factor solution, with the first factor being employee value. The eigenvalue for this factor is 7.38. The coefficient alpha for the eight items included in this factor is .83. The items are: (1) company's approach to employees, (2) commitment of employees, (3) way employees work together, (4) overall culture of the company, (5) family atmosphere, (6) company's values, (7) training our employees receive, and (8) hiring practices. The other factors which resulted from the survey, but which are not used for this study due to the nature of the research question, are management, the rewards system, product, and t echnology. Descriptive statistics for both measures of employee value are found in Table 1.

Pace or rate of growth

Change in sales from year end 1993 to year end 1995 is used as a measure of rate of change or pace. Sales data were gathered from COMPUSTAT.

Control Variables

Several control variables were used in the analysis. Those include organizational size, number of employees at the time of the IPO (logged to correct for skewness), risk (measured as the number of paragraphs in the risk section of the prospectus), stock price per share at the time of the IPO, net income at the time of the IPO, and industry (using the groupings recommended by the Small Business Administration).

Dependent Variable

Firm performance is measured as the percentage change in stock price from the time of the IPO to the end of 1995. In order to correct for skewness, the square root of this number is used in the analyses.

Data Analyses

Two separate regression analyses are run. The first includes the control variables, the measure of employee value taken from the prospectus, change in sales from years ending 1993 to 1995, and an interaction term with employee value and sales change. The second analysis is the same, except that the survey measure of employee value is used in the regression equation. The resulting sample size for the second analysis is smaller (250), as a result of missing data. The survey measure was obtained by averaging the responses of all individuals who responded within one firm. The correlation between the average response and the response from the highest ranking officer is .93.


Both analyses support hypothesis 1 in that there is a significant interaction between valuing employees and r ate of growth on stock price growth (see interaction graphs). The first analysis that used the measure of employee value obtained from the prospectus resulted in an R2 of .16 (F=4.92, p < .000) (See table 2). In order to interpret the interaction, I plotted the data based on splitting the sample into four categories (low and high on valuing employees and low and high on sales growth). Companies with high scores on value and sales growth have a 42% increase in stock price. Those low on both variables have an 8% increase in stock price. For those firms low on sales growth, valuing employees increases stock price growth from 8% to 27%. For those high on sales growth, valuing employees increases stock price growth from 39% to 42%. The more significant gain in performance is from employee value seems to be seen in the low sales growth firms.


Sales Growth

    low high


low 8% 39%
  high 27% 42%



Sales Growth

    low high


low 20% 13%
  high 36% 69%

The result of the second analysis (see table 2), using the measure of value obtained from the survey, is similar in that the interaction term is also significant. However, the pattern of results is different, and this analysis is supportive of hypothesis 1 in that the larger gain from valuing employees is evident in the high sales growth group of firms. The total R2 for this equation is .18 (F=3.08, significant at the .000 level). The analysis using the same four categories shows that companies claiming employees are highly valued and that have faster sales growth experience a change in stock price of 69%, while those low on both variables have a 20% increase in stock price. For those organizations low in sales growth, being low on value results in a 20% increase in stock price, while those high on value have a 36% increase in stock price. For those firms with high sales growth, the low value firms experience a 13% increase in stock price compared to the 69% growth rate for high value organizations. Thus, using the survey data, the higher gain is seen in the fast sales growth group of firms. However, in both samples, valuing employees leads to higher performance, whether sales growth is high or low.

Top of page | Chapter Listing | Return to 1997 Topical Index

1997 Babson College All Rights Reserved
Last Updated 06/01/98