Studies in High-Performance Human Resource Practices - Training and Development

In recent years, researchers have focused on demonstrating the value of human resource practices in improving organizational performance. Various labels are attached to the set of practices, but "high-performance human resource (HR) practices"seems most appropriate. A core part of high-performance human resource practice is a strong commitment to employee development, particularly of skills critical to the success of the organization. Thus, these studies provide useful evidence of development's value, although they do not isolate the effect of development alone.

Because these studies appear in academic journals, many practitioners are not aware of them. This section will provide key findings from these studies that are useful to HR practitioners. The primary advantage of the findings in this section is that they are based on rigorous research methodologies that are usually generalizable to some extent. All represent organization, or plant-level, data,as opposed to traditional program-level results assessment and are considered key studies in justifying the value of HR development.

Studies of a Broad Cross-Section of Companies
Huselid(1997) surveyed 3,452 firms representing all major industries in the United States. The final sample was of 968 firms. The survey consisted of thirteen high-performance work practices that were clustered into two groups:

Employee skills and organization structure (which contains development) and employee motivation. He then obtained financial data on each firm from publicly available sources to assess the impact of the high-performance practices on turnover, productivity (sales per employee) and financial performance (Tobin'sQ and gross rate of return on capital) while controlling for a number of factors such as firm size and industry.

Practical Results of Huselid

The results are encouraging. Clearly, development pays off in lower turnover,higher productivity, and improved financial performance.

Guzzo,Jette, and Katzell (1985) used a statistical technique called meta analysis to aggregate the results from ninety-eight studies where what we now call high-performance work practices were introduced into organizations. The practices they examined included recruitment and selection, training and instruction, appraisal and feedback, management by objectives, goal setting, work redesign, financial compensation, decision making techniques, sociotechnical interventions,supervisory methods, and work rescheduling. Three types of performance were examined: output, withdrawal from the organization, and disruptions in the organization (accidents, strikes, etc.). Thus, this is a very comprehensive analysis. The results are impressive:

• Over all the studies and intervention types, the average productivity of the worker receiving the interventions was 0.44 standard deviations above the average worker not receiving them. This is a very substantial difference.
• The effect was largest for disruption measures (effect = 0.82 std. dev.) and next largest for output (effect = 0.63 std. dev.).
• Training, goal setting, and sociotechnical systems design had the most powerful effects.
• For training alone, the average productivity for those receiving the interventions was 0.78 standard deviations above the average worker.
• The largest effect for training was 0.85 on the output measure, indicating that workers receiving training interventions performed 0.85 standard deviations above the worker not receiving training.

The National Organizations Study (NOS) is a special module of the General Social Survey that was conducted in 1991 with the support of the National Science Foundation(Kalleberg, Knoke, Marsden, and Spaeth 1996). The NOS used a special procedure to obtain a sample of 727 firms representative of U.S. work establishments. A variety of questions were asked of organizations about their HR practices and their performances relative to their competitors. Five performance scales were developed to measure: Products (product quality and development);employees (attract and retain essential employees); customer satisfaction,relations, and market factors (marketing, growth in sales, profitability, and market share).

Results showed that training was significantly correlated with all performance measures except customer satisfaction. The specific correlations are illustrated in Table below:

Correlation of Training and Perceived Performance Measures

Delaney and Huselid (1996) also used the NOS data set but analyzed the data a little differently. They constructed measures of nine different HR practices from the NOS data and tested the ability of these measures to predict two performance measures: perceived organizational performance and perceived performance in the product marketplace. The correlation of training alone with performance in the product marketplace was 0.19 but only 0.06 with organizational performance. Training and the other HR practices together explained approximately 18.5percent of the variance in organizational performance and 26.5 percent of the variance in product performance.

Correlations of Regression Analysis of Training with Perceived Performance

Measures

Bartel(1994) examined a cross-section of companies that had below average productivity in 1983. Her analysis showed that these firms chose to implement training programs in about half of their employee groups as a result of the lower productivity instead of implementing other HR interventions. The result was an18.86 percent productivity gain over a three-year period.

Studies of Manufacturing Plants
Youndt,Snell, Dean, and Lepak (1996) selected 512 manufacturing plants to study the effect of HR and manufacturing practices on plant performance.
Data obtained from the final sample of 100 plants included measures of operational performance,human resource practices (two parts—human capital enhancing and administrative), and four dimensions of manufacturing strategy.

Two key results emerged:

• Only 15 percent of the variation in performance was explained by manufacturing strategy alone. When HR systems were added, 36 percent of the variation in performance was explained, and only the human capital enhancing dimension of HR practices was statistically significant.
• The effects of human capital enhancing HR practices actually occurred in combination with quality oriented manufacturing strategies.

Their results clearly indicate that development interventions that enhance human capital are very important in conjunction with high-quality initiatives. Furthermore,human capital enhancing HR practices explained more of the variation in performance than did manufacturing strategy.

Lamand White (1998) also studied manufacturing firms from fourteen different industries. Using a survey, they assessed the "HR orientation" of each company,which they defined as "a systematic organizational effort to attract, retain,and develop competent and committed employees" (357). Using an interesting methodology, they asked HR executives to provide information about two competing firms rather than their own company. This data was analyzed for the effect on three measures of financial performance: return on assets, growth in sales, and growth in stock values while controlling for various extraneous factors.The results showed that HR orientation was important over and above the influence of control variables. Together with controls, HR orientation explained26 percent of the variability in return on assets; explained 25 percent of the growth in sales.

Arthur(1994,1992) developed a technique that classified steel "mini-mills" in the United States in two groups: those with a control oriented culture, and those with a commitment culture. The commitment culture was defined to include high levels of employee development along with participation. Results showed that those with commitment cultures had higher productivity (lower labor hours perton) and half the turnover rates of the control oriented cultures. Survival of Firms with IPOs

Welbourne and Andrews (1996) examined the prospectuses of 136 non financial IPO (initial public offering) firms to determine their emphasis on human capital/resources,including training. They then analyzed the data to determine the impact of emphasizing human capital on market value and five-year survival rates. After controlling for key extraneous variables, they found that the firms placing greater emphasis on human resources were more likely to survive for five years.In fact, valuing human resources along with rewards for performance correctly predicted 71 percent of the firms that survived for five years.

Retail Store Study
Russell,Terborg, and Powers (1985) examined a large international retail organization to assess the effects of training on store performance. In this organization, training programs for sales personnel were developed at the corporate level and distributed to each store. Store managers then made the decision about how many sales personnel to train. The average percent of sales personnel trained was31.5 percent, but ranged as high as 80 percent.

Store personnel were also asked to rate the degree of training emphasis in the store.Key results were:

• These two training variables explained 12 percent of the variability in store performance and 32 percent of the variability in store image, a key performance measure in retail.
• Each percentage point increase in people trained resulted in a sales gain of$37,258 for the average store. • Improvement of one-half standard deviation in percentage trained (about 16percentage points) increased store sales by$613,646.
• This was true even while controlling for the amount of support employees received.

Quality of Work-Life Companies

Lauand May (1998) conducted an interesting analysis to compare companies selected as the best companies to work for in the United States with the Standard and Poor's top one hundred companies (S&P 100). The first group they labeled"quality of work-life companies" because the selection criteria centered around employee satisfaction, retention, and growth opportunities.They selected four financial measures: sales growth, asset growth, return on assets, and return on equity.

The results are shown in Table below:

Comparison of Quality of Work Life and S&P Companies

The difference between the two groups was statistically significant for all except return on equity. Thus, companies with higher quality of work life grew faster and had a higher return on assets. One must be very careful to not conclude that quality of work life causes higher returns because the cause and effect relationships cannot be determined from this data. Nonetheless, the association between quality of work life and financial returns is striking.

Burke and Day (1986) used meta-analysis to aggregate the results of seventy studies examining the effectiveness of leadership or managerial training. Their classic article shows the effectiveness of managerial development interventions.

Their key results include:

• For studies using objective behavioral outcome measures (expertise): The average effect was to change behavior by 0.38 standard deviations, a very substantial change. Motivation training had the largest effect on behavior(effect = 0.81 std. dev.), human relations training having the next highest effect (effect = 0.41 std. dev.).
• For studies using objective performance results outcome measures: The average effect was to change results by 0.67 standard deviations, a very substantial change.

Human relations had the largest effect (effect = 1.04 std. dev.) with general management having substantial effect as well (effect = 0.53 std. dev.).

Conclusion

The high-performance work practices studies generally bundle training or development with other related HR practices so that in one sense they are limited in their ability to speak to the value of development interventions alone. However, they also represent the most rigorous studies documenting the value of key HR strategies that have development as a core component. Thus, we suggest they provide important evidence that supports development as a key to strategic advantage and organizational performance. The correlations and percent of variance explained in the above studies are considered substantial from a practical viewpoint and collectively lead us to conclude that Core Conclusion #2: Development should be a central part of HR strategy and HR strategy is a major determinant of organizational performance.