The Impact of Incentive Compensation on Labor Productivity in Family and Nonfamily Firms

Abstract

Family and nonfamily firms both must align owner and employee interests. However, family firms may experience lower labor productivity because of adverse selection problems from labor market sorting and attenuation. Incentive compensation reduces alignment of interest problems in family and nonfamily firms. Importantly, incentive compensation signals to potential employees that performance will be rewarded, which should improve the relative labor productivity in family firms by reducing adverse selection. Analysis of matched data on 216,768 firms supports our hypotheses, implying that incentive compensation has a broader impact on firm performance than commonly recognized in the family firm or human resource literatures.

Keywords

family firms incentive compensation nonfamily employees adverse selection labor productivity

Introduction

Measuring and rewarding employee performance in an equitable manner should improve labor productivity (De Kok, Uhlaner, & Thurik, 2006). However, such policies may be difficult to implement in family firms that pursue noneconomic goals in addition to economic goals (Chrisman, Chua, & Litz, 2004), have a propensity to engage in altruism toward family members (Schulze, Lubatkin, Dino, & Buchholtz, 2001), and exhibit bifurcation bias in favor of family employees at the expense of nonfamily employees (Verbeke & Kano, 2012).¹ For example, family owner-mangers might believe that using incentive compensation to attract the most able nonfamily employees will diminish their overall utility because fewer rewards will be available to distribute to family employees (Chrisman, Memili, & Misra, 2014; Chua, Chrisman, & Bergiel, 2009).

Recent research suggests family firms are less likely than nonfamily firms are to use incentives to compensate and motivate nonfamily employees (Memili, Misra, Chang, & Chrisman, 2013). However, some family firms are known for providing superior training, inclusiveness, job security, and employee support (Bammens, Notelaers, & Van Gils, 2015; Miller, Le Breton-Miller, & Scholnick, 2008), which may serve as effective substitutes for incentive compensation. Nevertheless, understanding the role of incentives in family firms is of theoretical and practical importance, as it deals with how employees are motivated (Chrisman et al., 2004; Corbetta & Salvato, 2004).²

The direct impact of incentive compensation on firm performance in family as well as nonfamily firms is perceived to be positive, and this has largely been borne out by empirical research (e.g., Chrisman, Chua, Kellermanns, & Chang, 2007; Long & Fang, 2015; Michiels, Voordeckers, Lybaert, & Steijvers, 2013). However, in addition to providing incentives to maximize employees’ effort, firms also need to attract high-quality employees since performance is also dependent on employees’ ability. Unfortunately, attracting able employees is more difficult when family firm owners pursue noneconomic goals and favor family employees over nonfamily employees (e.g., Carney, 2005; Chua et al., 2009; Verbeke & Kano, 2012; Schulze et al., 2001), because potential nonfamily employees are apt to presume they will not be adequately rewarded for their contributions to the firm (Chua et al., 2009). Since the most able workers have the most options, they are likely to sort themselves according to their preferences to work in nonfamily firms where career prospects and compensation levels are perceived to be superior (Block, Fisch, Lau, Obschonka, & Presse, 2016; Chrisman et al., 2014). To the extent that such sorting occurs, the labor market from which family firms can draw employees becomes attenuated, leading to an adverse selection problem.

Adverse selection describes the risk to boundedly rational firm owners and managers associated with hiring individuals who are less able or motivated than expected. Adverse selection occurs when potential employees opportunistically misrepresent their qualifications (Eisenhardt, 1989) and/or engage in self-sorting according to the perceived self-interest associated with working at a firm with specific attributes, in this case, income earning and career enhancing opportunities (Chrisman et al., 2014). The adverse selection problem from either opportunistic misrepresentation or sorting based on employment preferences exists for all firms. However, when higher quality performers systematically sort themselves out of the labor market for a particular type of firm, the labor market for such firms becomes attenuated. Labor market attenuation owing to sorting decreases the average quality of potential applicants in a way that is not readily observable, creating information asymmetries for boundedly rational employers concerning the composition of the labor market, and making the expected quality of job candidates more difficult to estimate (Akerlof, 1970). Herein, we argue that the labor market sorting and attenuation problem, is more acute for family firms and that this problem will put family firms at a performance disadvantage vis-à-vis nonfamily firms.

However, incentive compensation can act as a credible signal that rewards will be commensurate with performance (Spence, 1973). This makes employment in family firms a more attractive option for higher quality workers, reducing labor market sorting and attenuation since the average quality of the pool of potential candidates available for family firms to hire will approach the average quality of workers in the general labor market. This makes worker quality easier to predict, which increases the probability of hiring high-quality workers. To the extent that the constrains family owners face in hiring workers of higher quality are alleviated, the labor productivity of their firms should increase relative to nonfamily firms, which are generally not expected to suffer from the problems of labor market sorting and attenuation discussed above.

We test these arguments using matched data on 216,768 companies from the 2007 Survey of Business Owners. Specifically, we investigate whether incentive compensation influences the labor productivity of family firms relative to nonfamily firms. We contribute to the literature by providing evidence that incentive compensation has a greater relative impact on labor productivity in family firms than in nonfamily firms. Since the alignment of interest problem that incentive compensation helps to solve exists in both family and nonfamily firms, we attribute this finding to a reduction in labor market attenuation caused by sorting, a problem that is expected to be more severe in family firms. In other words, besides aligning the interests of owners and employees to increase the motivation and effort of the latter, incentive compensation is expected to increase the relative attractiveness of family firms to workers with greater abilities, thereby raising the average quality of the pool of potential employees and, ultimately, firm performance.

We also contribute to the literature by illustrating that the adverse selection and alignment of interest problems in family firms might have causes (e.g., limited opportunities for nonfamily employees), consequences (e.g., higher costs, lower labor productivity), and solutions (e.g., incentive compensation) in common. Furthermore, since ability and effort tend to be positively related (cf., Bandura, 1977, 1982; Chua et al., 2009), improving the ability of the workforce should lead to greater effort. Thus, our findings suggest incentive compensation can potentially improve the productive ability and productive effort of the workforce in family firms.

Theoretical Background and Hypotheses

Previous research indicates that incentive compensation helps align the interests of owners and employees and is associated with higher productivity (FitzRoy & Kraft, 1987; Florkowski, 1987; Kruse, Blasi, & Freeman, 2012; Kruse, Freeman, & Blasi, 2010; Meade, 1972; Steinherr, 1977). When workers’ pay is tied to firm profits, they should begin to act more like owners, sharing a common interest in the prosperity of the firm. Firms that offer superior compensation packages are also more likely to attract and retain high-quality employees (Hancock, Allen, Bosco, McDaniel, & Pierce, 2013; Long & Fang, 2015), which can further increase labor productivity. Importantly, the attractiveness of an employer that offers incentive compensation should be greater for high-quality workers than for low-quality workers. This is because the former (latter) are more (less) likely to believe they have the ability to positively affect firm performance and are more (less) willing to expend the effort necessary to do so (Bandura, 1977, 1982; Chua et al., 2009; Long, 2000).

Incentive compensation increases employee cooperation and coordination as well as motivation (Blinder, 2011; Kruse et al., 2012). By aligning the goals of owners and employees, incentive compensation encourages mutual monitoring efforts and the exertion of peer pressure on free-riding coworkers (Kruse et al., 2012). Firms that offer incentive compensation are also more inclined to pay higher base wages and invest in training and development, which should further increase employee effort, coordination, cooperation, and, consequently, firm performance (Chandler & McEvoy, 2000; Hart & Huebler, 1991; Long & Fang, 2015; Singh, 2004). However, even firms with lower ability to pay high fixed wages can benefit from offering incentive compensation as it allows them to increase total employee pay, thereby attracting employees that are more qualified (Long, 2000). In this sense, we assume that information about specific firms and classes of firms will be disseminated through formal and informal networks, resulting in the sorting or matching of potential employees and employers according to their attributes (Chrisman et al., 2014). Thus, our baseline hypothesis is the following:

Hypothesis 1: Firms with incentive compensation programs will have higher labor productivity than firms without incentive compensation programs.

Incentive Compensation in Family and Nonfamily Firms

The problem of aligning the interests of employees with those of owners is universal. However, in family firms, which are sometimes known to channel firm resources toward family members and to favor family employees over nonfamily employees (Chua et al., 2009; Schulze et al., 2001; Verbeke & Kano, 2012), the difficulty in obtaining the commitment of nonfamily employees to the firm may be even greater (Barnett & Kellermanns, 2006). As Dyer (2006, p. 264) states, “Nonfamily employees are [often] treated as ‘second-class citizens’ and . . . such an adversarial relationship between an owning family and nonfamily employees often results in low employee morale and low productivity.” A desire to exercise significant decision-making authority and to maintain ownership and control within the family across generations can limit opportunities for advancement and financial rewards for nonfamily employees (Memili et al., 2013). As suggested above, this means that the labor market from which family firms can draw personnel may be attenuated because the best workers with the most employment options may opt to work in nonfamily firms where opportunities for immediate and long-term rewards are not a function of membership in the owning family (Chrisman et al., 2014). The family’s pursuit of noneconomic goals, as well as economic goals, makes recruiting highly qualified employees even more difficult because such goals tend to lead to idiosyncratic strategies with which potential nonfamily employees are both unlikely to be familiar and unlikely to benefit (Chua et al., 2009). Thus, because of sorting and attenuation of the labor market, family firms must deal with an adverse selection problem that is more acute than for nonfamily firms that are not perceived to exhibit systematic bias against nonfamily employees.

Furthermore, since ability and effort tend to be correlated (cf., Bandura, 1977, 1982), the negative impact on performance of a reduced capacity to attract high-quality workers will likely be intensified for family firms (Chua et al., 2009), because the costs of producing a given level of output will be greater. To maximize their utility, workers that are less able are likely to expend less effort than workers that are more able, which will cause firm performance to suffer. Family firms may thus end up with less able employees who also put forth less effort.

Finally, opting to employ family members rather than nonfamily members may help avoid the problems noted above and contribute to the achievement of family owners’ noneconomic goals (Chrisman et al., 2014). However, family employees selected based on relationships rather than merit, tend to have lower ability than nonfamily employees selected based on merit (Cucculelli, Mannarino, Pupo, & Ricotta, 2014; Karra, Tracey, & Phillips, 2006; Schulze et al., 2001). For example, Barth, Gulbrandsen, and Schøne (2005) found that among Norwegian firms with 100 employees or less, productivity was 12.6% lower among family firms managed by a family member than among family firms managed by a nonfamily member.

On the other hand, family firms are considered attractive employers for a number of reasons. As noted earlier, family owner-managers are often known for their nurturing and support of employees, including nonfamily employees (Bammens et al., 2015). Along the same lines, family firms are associated with providing greater job security and more flexible work practices (Bach & Serrano-Velarde, 2015), which facilitates their ability to attract employees who value work environments that provide amenities such as security, stability, and benevolence (Block et al., 2016; Hauswald, Hack, Kellermanns, & Patzelt, 2015; Miller et al., 2008).

How these positives and negatives of family firms combine to influence firm performance is not completely clear. Indeed, the literature on the relationship between family ownership and labor productivity has been equivocal with some studies of nonlisted firms finding the relationship to be negative (e.g., Morikawa, 2013), some finding the relationship to be positive (e.g., Barbera & Moores, 2013), and some studies finding the relationship to be contingent on the management regime (Barth et al., 2005). Overall, however, we take the position that the negatives will usually outweigh the positives because flexible work practices, security, and a benevolent work environment appeal to everyone and require no special skills to attain, whereas ability and effort are necessary to take full advantage of income-earning and career advancement opportunities (Chrisman et al., 2014). Furthermore, the career uncertainties associated with the family firm option are likely to weigh more heavily on prospective recruits with the highest career aspirations and abilities. Although family firms may be more nurturing, the perception, if not the reality, that some family owner-managers will exhibit favoritism toward family members might increase the probability that prospective employees with high potential will seek employment in nonfamily firms where the relationship between performance and rewards is seen as more predictable. Again, because labor market attenuation reduces the pool of job candidates, and although some family firms might be able to secure suitable personnel, this sorting affect is likely to place family firms at an information disadvantage in selecting individuals who will contribute to firm performance.

Given a choice between high short- and long-term nonpecuniary inducements and high short- and long-term pecuniary inducements, we reason that the latter will disproportionately attract higher quality workers. Indeed, research suggests family firms tend to pay less than nonfamily firms (Bach & Serrano-Velarde, 2015) and struggle to attract highly qualified personnel (Block et al., 2016). Thus, because labor productivity depends on the ability, as well as the effort of employees, which are interrelated, we propose the following:

Hypothesis 2: Family firms will have lower labor productivity than nonfamily firms.

Incentive Compensation and Labor Productivity in Family and Nonfamily Firms

As noted earlier, the owners of family firms are often reluctant to provide incentive compensation to nonfamily employees because doing so is perceived as an impediment to the achievement of noneconomic goals such as the preservation of family control and the provision of benefits to family members (Chrisman et al., 2014; Memili et al., 2013). Furthermore, family firms tend to pay lower wages than do nonfamily firms (Bach & Serrano-Velarde, 2015). Relative to nonfamily firms, family firms are also more likely to implement idiosyncratic strategies to achieve noneconomic goals (Gomez-Mejia, Haynes, Núñez-Nickel, Jacobson, & Moyano-Fuentes, 2007), which places nonfamily employees at a disadvantage because their prior experience will not be as useful in a context where economic performance is a less important driver of firm behavior. Likewise, the pursuit of idiosyncratic strategies to achieve noneconomic goals could make work experience in a family firm more difficult to transfer should a nonfamily employee decide to leave (Chrisman et al., 2014; Chua et al., 2009).

Family firms that institute incentive compensation programs, however, send a signal of commitment to superior economic performance, which suggests they follow a strategy that is more in line with the prior experience of the skilled workers they desire to attract. Incentive compensation programs also specifically suggest greater rewards for high-performing nonfamily employees and an absence, or at least an abridgement, of biases in favor of family employees. In other words, the use of incentive compensation signals that noneconomic goals are relatively less important than would otherwise be anticipated. We expect this to be true because the cost of this form of signaling should be negatively correlated with the relative importance of economic versus noneconomic goals (see Spence, 1973). In other words, the more noneconomic goals dominate decision making (and hence the greater importance of family control and greater propensity toward idiosyncratic strategies and family biases), the more utility family firms have to give up to institute profit-sharing, making such plans less attractive and less likely to be implemented (Chrisman et al., 2014). On the other hand, since nonfamily firms do not face an attenuated labor market caused by sorting, they are unlikely to have the same problems in attracting high-quality employees or in estimating the quality of potential employees that are available to hire. Consequently, the benefits of incentive compensation for nonfamily firms will not be as great as it is for family firms.

Thus, incentive compensation programs send a signal about the priorities of the family firms that implement such programs, which should increase the odds that potential nonfamily employees will view them as attractive places to work. Since higher quality workers are the most likely to benefit from incentive compensation programs, the probability of them applying for positions in family firms should increase. While this does not mean that family firms will necessarily be considered more desirable potential employers than nonfamily firms by the most able job candidates (e.g., upward mobility might still be limited), the use of incentive compensation is expected to narrow the gap in labor productivity hypothesized above. Thus, for family firms, incentive compensation should reduce the adverse selection problem caused by labor market sorting and attenuation, which is a source of their managerial capacity constraints discussed in the literature (e.g., Carney, 2005).

Hypothesis 3: Incentive compensation will increase the labor productivity of family firms more than it will increase the labor productivity of nonfamily firms.

Data and Method

To test the hypotheses, we use data from the 2007 Survey of Business Owners (SBO) obtained from the Public Use Microdata Sample (PUMS) released by the U.S. Census Bureau in August 2012. The sampling frame includes all nonagriculture businesses that were in operation during 2007, had revenues of at least $1,000 during the year, and filed tax returns with the Internal Revenue Service using at least one of the following forms: 1040 Schedule C, Form 1065, Corporation Tax Form 1120, Form 941, or Form 944. Overall, 62% of the firms responded with complete information to the survey. Further details on the data collection are available on PUMS SBO 2007 website.³

The initial sample includes 2,165,680 firms. We first drop firms the owners reported were no longer in operation (880,006 firms). Out of the remaining sample, we drop firms that did not have paid employees (612,160 firms). To ensure greater comparability among the firms in the sample and to avoid possible ambiguities among the classification of family firms, we also excluded single owner (lone founder) firms (296,118 firms). We did include husband–wife firms as they are construed as family firms. However, the inferences were not different by excluding husband–wife firms. We also drop observations with missing values on the dependent, independent, or control variables (44,460 firms). This led to a sample of 332,936 firms, of which 224,552 were family firms and 108,384 were nonfamily firms.

Owing to our application of the propensity score matching method of analysis, the sample was further reduced to 216,768 (108,384 family firms and 108,384 nonfamily firms). Matched pair sampling is particularly salient in the current context for theoretical and empirical reasons. Theoretically, propensity score matching ensures firms in the sample are comparable on their key characteristics, and such comparability allows for a more robust test of the proposed differences in labor productivity among family and nonfamily firms. Empirically, in large census-type data sets, using the entire sample could lead to significant bias due to unobserved heterogeneity among the firms. As census data generally includes stratified samples, sample weights inherent in such samples could also bias the estimates.⁴ By matching firms on key characteristics, unobserved heterogeneity among the firms and the effects of sampling bias are reduced (Dehejia & Wahba, 2002; Peikes, Moreno, & Orzol, 2008). Including matched firms also provides more conservative estimates and lowers the chances of Type 1 error, by controlling for “confounding factors, such as size and industry” (Hambrick & D’Aveni, 1988, p. 7).

From the causality viewpoint, using cross-sectional, census-type data makes obtaining a set of instruments to test for endogeneity more difficult (Caliendo & Kopeinig, 2008; Rubin, 2006). Likewise, the presence of substantial heterogeneity in the sample leads to correlation between the error term and covariates that vary in strength and increase possible reverse causality across different intertemporal ranges in the predictor variable. Propensity score matching at least partially mitigates these potential drawbacks by identifying firms that are comparable on a multitude of firm-related characteristics and assessing performance gaps based on incentive compensation “treatments” (Caliendo & Kopeinig, 2008).

For the analyses, we use matching without replacement based on the following variables: incentive compensation, firm size, firm age, employee benefits (health insurance, retirement programs, and paid holidays), nature of business (seasonal business, franchise, and home-based business), industry, and geographic sector (state). Matching without replacement, which avoids double counting and provides one-to-one comparisons of family and nonfamily firms, is a stronger method.⁵ Nevertheless, matching with replacement yielded similar results. In addition to using propensity score matching, we use 2SLS regression to control for endogeneity and perform a variety of additional analyses to ensure the robustness of our results.

Variables

Dependent Variable

We use labor productivity as the dependent variable, measured as the ratio of firm revenues to number of employees. We use the log value in our analysis because labor productivity is not normally distributed. This measure has been used in prior studies of small and large firms (e.g., Kruse et al., 2012) and is appropriate for this study because higher labor productivity is dependent on worker ability and effort. Furthermore, higher labor productivity should lead to higher profits (i.e., revenues less costs). All else held equal, an increase in sales per employee suggests that more revenue is being generated for a given level of labor cost.

Independent Variables

To operationalize the family business variable, we used the question in the SBO 2007 survey that asks subjects whether two or more members of the same family own the majority of this business, where family refers to spouse, parent/guardians, children, siblings, or close relatives. Positive responses were coded as one (yes = 1) and negative responses were coded as zero (no = 0). While multidimensional measures of family firms would be preferable, the use of a dummy variable is frequently found in the literature (e.g., Anderson, Duru, & Reeb, 2012; Boling, Pieper, & Covin, 2016; Chrisman et al., 2014; Gomez-Mejia et al., 2007). Indeed, recent studies using multiple indicators of the family firm construct find they are strongly intercorrelated (Gomez-Mejia et al., 2014), suggesting our measure is acceptable.

Consistent with prior studies (Kruse et al., 2012; Long, 2000), to measure incentive compensation, we used responses to a question asking whether the firm provided employees with profit sharing and/or stock options (yes = 1, no = 0).

Control Variables

Smaller firms may face challenges in implementing incentive compensation programs (Way, 2002) and in realizing higher labor productivity. Furthermore, smaller firms are less likely to employ individuals who are not part of the owning family (Fang, Randolph, Memili, & Chrisman, 2016). Therefore, we control for firm size using the log of total employees as a proxy. As firm age affects labor productivity (Way, 2002), we control for different age categories, with “before 1980” as the reference category.

The average wages paid by a firm is often correlated with the use of incentive compensation (Long & Fang, 2015) and can act as an inducement for higher ability job candidates to work in a family firm (Chrisman et al., 2014). Therefore, the log of payroll per employee is used as a control for our analysis. We also use a benefits index as a control. That index includes the sum of each respondent’s answer to questions concerning whether the firm offers health insurance, retirement benefits, and paid holidays to employees (yes = 1, no = 0 for each). Since benefit packages are, in effect, substitutes for pay, they could influence labor productivity for the same reasons that incentive compensation plans (or wage levels) do.

To control for the nature of the business, we include whether the firm is a seasonal, franchise, or home-based business. We also controlled for the state where the firms are headquartered, with Alabama as the reference state, and industry sector, with forestry and wood products as the reference two-digit SIC code industry sector. The latter measure is important because previous research has indicated that the willingness of family firms to use nonfamily employees varies by industry (Fang, Memili, Chrisman, & Penney, 2016).

Results

Table 1 presents variable means, standard deviations, and correlations. To assess average treatment effects of incentive compensation using propensity score matching (Caliendo & Kopeinig, 2008), we employ the psmatch2 package in Stata 11.2. The distribution of propensity scores was uniform and balanced, and there was no significant bias in the distribution of covariates across matched (firms with and without incentive compensation plans) and unmatched cases (firms without incentive compensation plans that were not comparable with firms with incentive compensation plans).

Table 1.

Sample Descriptives.

	M	SD	Min	Max	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21
1. Log of productivity	4.8412	1.2925	0	11.8845	1
2. Log of total employment	2.9475	1.3650	0.6931	10.4631	−0.0599*	1
3. Log of pay per employment	3.5869	0.8386	0	9.9035	0.5087*	0.0269*	1
4. Benefits index	2.1908	0.9794	0	3	0.2067*	0.4108*	0.4199*	1
5. Health insurance	0.8038	0.3971	0	1	0.1713*	0.3467*	0.3539*	0.7776*	1
6. Retirement benefits	0.5551	0.4970	0	1	0.1627*	0.3024*	0.3193*	0.8028*	0.4113*	1
7. Holidays	0.8318	0.3740	0	1	0.1431*	0.3058*	0.2995*	0.7264*	0.4281*	0.3368*	1
8. Seasonal business	0.0116	0.1069	0	1	−0.0327*	−0.0308*	−0.0562*	−0.0653*	−0.0544*	−0.0318*	−0.071*	1
9. Franchise	0.0443	0.2058	0	1	−0.1076*	0.0745*	−0.1609*	−0.0568*	−0.0355*	−0.0647*	−0.0251*	0.0166*	1
10. Home-based business	0.0407	0.1976	0	1	−0.0504*	−0.2089*	−0.0762*	−0.1962*	−0.1505*	−0.1025*	−0.2179*	0.0275*	−0.0333*	1
11. Established before 1980	0.2590	0.4381	0	1	0.0608*	0.2463*	0.1251*	0.2649*	0.1964*	0.2353*	0.1725*	0.0034	−0.0683*	−0.1004*	1
12. Established between 1980 and 1989	0.2061	0.4045	0	1	0.0202*	0.068*	0.0652*	0.1286*	0.104*	0.1006*	0.0925*	−0.0074*	−0.0186*	−0.0573*	−0.3012*	1
13. Established between 1990 and 1999	0.2605	0.4389	0	1	−0.001*	−0.022*	0.0074*	0.0107*	0.0256*	−0.0107*	0.0151*	−0.0028	0.0119*	−0.0039	−0.3509*	−0.3024*	1
14. Established between 2000 and 2002	0.1006	0.3008	0	1	−0.0103*	−0.0814*	−0.0429*	−0.1118*	−0.0798*	−0.1046*	−0.0691*	−0.0044*	0.0164*	0.0483*	−0.1978*	−0.1704*	−0.1986*	1
15. Established in 2003	0.0364	0.1872	0	1	−0.0205*	−0.0686*	−0.0491*	−0.0996*	−0.0783*	−0.0884*	−0.0601*	−0.0026	0.0192*	0.0394*	−0.1149*	−0.099*	−0.1153*	−0.065*	1
16. Established in 2004	0.0421	0.2007	0	1	−0.028*	−0.0954*	−0.0744*	−0.1439*	−0.1133*	−0.1188*	−0.0986*	0.0077*	0.0257*	0.054*	−0.1239*	−0.1068*	−0.1244*	−0.0701*	−0.0407*	1
17. Established in 2005	0.0407	0.1976	0	1	−0.0342*	−0.1181*	−0.0901*	−0.1764*	−0.151*	−0.1328*	−0.1252*	0.0046*	0.0373*	0.0709*	−0.1218*	−0.1049*	−0.1223*	−0.0689*	−0.04*	−0.0432*	1
18. Established in 2006	0.0364	0.1873	0	1	−0.0503*	−0.1422*	−0.1015*	−0.1957*	−0.1649*	−0.1438*	−0.1463*	0.0071*	0.0392*	0.0737*	−0.115*	−0.0991*	−0.1154*	−0.065*	−0.0378*	−0.0407*	−0.04*	1
19. Established in 2007	0.0182	0.1336	0	1	−0.0422*	−0.1443*	−0.0791*	−0.1502*	−0.1291*	−0.105*	−0.1167*	0.0056*	0.0282*	0.0621*	−0.0805*	−0.0693*	−0.0808*	−0.0455*	−0.0264*	−0.0285*	−0.028*	−0.0265*	1
20. Incentive compensation	0.2113	0.4083	0	1	0.1392*	0.1706*	0.2216*	0.2678*	0.1998*	0.2333*	0.179*	−0.0194*	−0.0607*	−0.051*	0.2321*	0.0229*	−0.0676*	−0.0759*	−0.0522*	−0.0652*	−0.0735*	−0.074*	−0.0537*	1
21. Family Business	0.5000	0.5000	0	1	−0.0616*	0.0365*	−0.1122*	0.0188*	0.0172*	0.0103*	0.0172*	−0.0018	0.0097*	−0.0442*	−0.0025	0.0141*	−0.003	−0.0093*	−0.0046*	−0.0024	0.0012	0.0019	0.0021	0.0017	1

Note. All estimates with * are significant at .05 or below (two-tailed test).

Table 2 provides a matrix comparing the family and nonfamily firms before and after the matching process. Based on the raw sample, in comparison to nonfamily firms, family firms were on average much less likely to offer incentive compensation (12.8% vs. 21.1%), health insurance (64.6% vs. 79.7%), retirement benefits (40.7% vs. 55.0%), or paid holidays (69.7% vs. 82.5%). Family firms, compared with nonfamily firms, also tend to have fewer employees (37 employees vs. 52 employees), pay less ($37,694 vs. $57,383), and were more likely to have started before 1980. Finally, the family firms in the sample were more likely to be in seasonal, franchise, and/or home-based businesses. Of course, after matching there was a much closer convergence between the two groups of firms on all these characteristics.

Table 2.

Family Versus Nonfamily Firm Comparison.

	Pre-matching				Post-matching
	Family business; N = 224,552		Nonfamily business; N = 108,384		Family business; N = 108,384		Nonfamily Business; N = 108,384
	M	SD	M	SD	M	SD	M	SD
Incentive compensation	0.1276	0.3337	0.2106	0.4078	0.2120	0.4088	0.2106	0.4078
Total employment	37.3031	140.2154	51.7646	247.5516	59.3391	195.7503	51.7646	247.5516
Log of total employment	2.5369	1.3332	2.8976	1.3309	2.9973	1.3965	2.8976	1.3309
Payroll per employee	37.6946	87.0327	57.3838	142.9274	44.0281	77.4907	57.3838	142.9274
Log of payroll per employee	3.3041	0.8691	3.6811	0.8622	3.4928	0.8033	3.6811	0.8622
Health insurance	0.6463	0.4781	0.7969	0.4023	0.8106	0.3918	0.7969	0.4023
Retirement benefits	0.4067	0.4912	0.5500	0.4975	0.5603	0.4964	0.5500	0.4975
Holidays	0.6969	0.4596	0.8254	0.3796	0.8383	0.3682	0.8254	0.3796
Seasonal business	0.0200	0.1401	0.0117	0.1077	0.0114	0.1060	0.0117	0.1077
Franchise	0.0605	0.2384	0.0423	0.2013	0.0463	0.2102	0.0423	0.2013
Home-based	0.1546	0.3616	0.0495	0.2168	0.0320	0.1759	0.0495	0.2168
Established before 1980	0.3614	0.4804	0.2601	0.4387	0.2579	0.4375	0.2601	0.4387
Established between 1980 and 1989	0.1993	0.3994	0.2004	0.4003	0.2118	0.4086	0.2004	0.4003
Established between 1990 and 1999	0.2280	0.4195	0.2618	0.4396	0.2592	0.4382	0.2618	0.4396
Established between 2000 and 2002	0.0810	0.2728	0.1034	0.3045	0.0978	0.2971	0.1034	0.3045
Established in 2003	0.0285	0.1664	0.0372	0.1893	0.0355	0.1850	0.0372	0.1893
Established in 2004	0.0318	0.1755	0.0426	0.2018	0.0416	0.1996	0.0426	0.2018
Established in 2005	0.0305	0.1719	0.0405	0.1970	0.0409	0.1981	0.0405	0.1970
Established in 2006	0.0265	0.1606	0.0361	0.1865	0.0368	0.1882	0.0361	0.1865
Established in 2007	0.0131	0.1139	0.0179	0.1326	0.0185	0.1346	0.0179	0.1326

Note. State and Sector dummies included in the matching.

Table 3 presents the means and standard deviations of labor productivity for family versus nonfamily firms, firms with and without incentive compensation plans, and the four respective subcategories. The labor productivity of firms offering incentive compensation was 9.3% higher than that of firms that did not offer incentive compensation. Overall, family firms had labor productivity that was 3.3% lower than in nonfamily firms. Although family firms with incentive compensation plans also had lower labor productivity than comparable nonfamily firms, the absolute gap in labor productivity was only −0.5%, a reduction in the gap of approximately 85% in relative terms. In addition, family firms with incentive compensation plans had labor productivity that was 11.3% higher than that of family firms without incentive compensation plans and 6.8% higher than that of nonfamily firms without incentive compensation plans. These results are consistent with our theoretical discussion and hypotheses.

Table 3.

Productivity Matrix of Firms in Matched Sample.

	Firms with incentive compensation plans (C)	Firms without incentive compensation plans (D)	Total
Family firms (A)	5.1765 (1.3504)	4.6500 (1.2744)	4.7616 (1.3087)
Family firms (A)	N = 22,981	N = 85,403	N = 108,384
Nonfamily firms (B)	5.2011 (1.2478)	4.8460 (1.2669)	4.9208 (1.2712)
Nonfamily firms (B)	N = 22,829	N = 85,555	N = 108,384
Total	5.1887 (1.3003)	4.7481 (1.2745)
Total	N = 45,810	N = 170,958

Note. A total of 108,384 family business matched with 108,384 nonfamily business firms. The matching variables are incentive compensation, log of total employment, health insurance, retirement, holidays, seasonal business, franchise, home based, year of establishment, state, and industry sector. Standard deviations are provided within parenthesis.

Results: 2SLS Model

To test the hypotheses, we use a 2SLS model using the sample of matched firms. Although we do not expect reverse causality from labor productivity to incentive compensation plans, there is a potential for endogeneity in the proposed specification stemming from unobservables in the error term that influence both the decision to use incentive compensation by family firms and the expected gains in labor productivity. Therefore, controlling for endogeneity could help derive inferences that are more robust. We use two instruments.

The first instrument is whether the firm is located in a right-to-work state. Right-to-work laws exist in 25 states in the United States and primarily aim to lower the influence of unions in a workplace by not requiring employees to pay union dues or be a part of a union. In states with right-to-work laws, firms have greater discretion in their compensation policies as well as pensions and health insurance (Moore & Newman, 1985; Stevans, 2009). Right-to-work laws should increase competitiveness and decrease protections for less productive workers. They may therefore enable and encourage employers to provide incentive compensation plans. However, while right-to-work laws may influence labor costs and managerial discretion, they would not appear to influence labor productivity directly.

The second instrument is the state minimum wage in 2007 adjusted for the cost of living in year 2000 real dollars. As not all costs of higher minimum wages could be transmitted to customers, businesses would bear part of the burden of higher wages. Higher minimum wages might influence the adoption of incentive compensation plans that motivate employees to increase labor productivity. Indeed, as noted above, incentive compensation plans and wages appear to be positively correlated (Long & Fang, 2015). However, as minimum wage legislation is pursued by external stakeholders to increase pay for workers employed in entry-level positions (Zavodny, 2000), and has varied over time across and within states, it, in itself, is unlikely to directly motivate workers to increase labor productivity.

The final instruments are created by separately multiplying these variables by the interaction of the family business and incentive compensation variables. In the first stage of the 2SLS model, the outcome variable is therefore family businesses using incentive compensation (Model 1, Table 4). The instruments were significantly related to the endogenous variable, family firms using incentive compensation. The Angrist–Pischke F test indicated that weak instrument bias does not exist (F = 1551.87). The Kleibergen–Paap LM chi-squared test rejects the null hypothesis of under-identification (χ² = 30.078). These tests satisfy the rank condition and provide support of instrument validity.

Table 4.

Endogeneity Analysis for 2SLS Tests.

	First stage (Model 1)
	Family Business × Incentive compensation
Family business	−0.0018 (0.0016)
Incentive compensation	0.0173** (0.0045)
Log of total employment	−0.0000 (0.0002)
Log of payroll per employee	−0.0010** (0.0003)
Log of payroll per employee × Family business	0.0000 (0.0006)
Benefits index	−0.0017** (0.0004)
Benefits index × Family business	0.0041** (0.0010)
Seasonal business	0.0013 (0.0009)
Franchise	0.0010⁺ (0.0005)
Homebased	0.0001 (0.0003)
Established between 1980 and 1989	−0.0040** (0.0012)
Established between 1990 and 1999	−0.0041** (0.0012)
Established between 2000 and 2002	−0.0039** (0.0012)
Established in 2003	−0.0035** (0.0011)
Established in 2004	−0.0037** (0.0012)
Established in 2005	−0.0036** (0.0011)
Established in 2006	−0.0034** (0.0011)
Established in 2007	−0.0033** (0.0011)
Industry dummies	Included
State dummies	Included
Instruments
Right to work state × Family business × Incentive compensation	0.2089** (0.0355)
Minimum wage × Family business × Incentive compensation	0.1412** (0.0031)
Intercept	0.0073** (0.0024)
Observations	216,768
R ²	0.986

Note. Robust standard errors clustered by state in parentheses; based on matching without replacement.

p < .10. *p < .05. **p < .01.

Hypothesis Tests

As shown in Table 5, we use the matched pair sample and the predicted values from Model 1 to test the hypotheses. First, as shown in Model 2, the incentive compensation variable is positive and significant (p < .01) and the family business variable is negative and significant (p < .01) in support of Hypothesis 1 and Hypothesis 2, respectively. Hypothesis 3 is also supported. The interaction between the incentive compensation and family business variables is positive and significantly related to labor productivity (p < .01).

Table 5.

2SLS Regression Analyses (Second Stage Tests).

	(Model 2)	(Model 3)	(Model 4)	(Model 5)
	lnprod	lnprod	lnprod	lnprod
Family business	−0.0248** (0.00400)	−0.160** (0.0189)	−0.0628** (0.00853)	−0.1608** (0.0185)
Incentive compensation	0.0567** (0.00894)	0.0775** (0.00718)	0.0772** (0.00715)	0.0665** (0.0090)
Incentive compensation × Family business	0.0407** (0.00979)			0.0223* (0.0104)
Log of payroll per employment	0.822** (0.00400)	0.803** (0.00532)	0.821** (0.00404)	0.8053** (0.0056)
Log of payroll per employment × Family business		0.0403** (0.00517)		0.0349** (0.0059)
Benefits index	0.0280** (0.00264)	0.0281** (0.00271)	0.0181** (0.00331)	0.0251** (0.0035)
Benefits index × Family business			0.0213** (0.00363)	0.0069 (0.0044)
Log of total employment	−0.111** (0.00334)	−0.111** (0.00337)	−0.111** (0.00338)	−0.1110** (0.0033)
Seasonal business	−0.00516 (0.0172)	−0.00446 (0.0174)	−0.00523 (0.0174)	−0.0049 (0.0172)
Franchise	0.102** (0.0101)	0.103** (0.0102)	0.102** (0.0102)	0.1028** (0.0101)
Homebased	−0.102** (0.00916)	−0.105** (0.00933)	−0.106** (0.00928)	−0.1068** (0.0091)
Established between 1980 and 1989	−0.0224** (0.00552)	−0.0228** (0.00551)	−0.0234** (0.00551)	−0.0217** (0.0055)
Established between 1990 and 1999	−0.00160 (0.00560)	−0.00130 (0.00564)	−0.00194 (0.00566)	0.0000 (0.0056)
Established between 2000 and 2002	0.0335** (0.00927)	0.0344** (0.00927)	0.0342** (0.00932)	0.0360** (0.0092)
Established in 2003	0.000643 (0.0109)	0.00227 (0.0112)	0.00188 (0.0114)	0.0039 (0.0110)
Established in 2004	0.0218* (0.00983)	0.0241* (0.00993)	0.0235* (0.0101)	0.0260** (0.0099)
Established in 2005	0.0185 (0.0116)	0.0209⁺ (0.0116)	0.0204⁺ (0.0117)	0.0229* (0.0114)
Established in 2006	−0.0630** (0.0113)	−0.0603** (0.0112)	−0.0608** (0.0114)	−0.0582** (0.0111)
Established in 2007	−0.119** (0.0169)	−0.115** (0.0172)	−0.117** (0.0175)	−0.1126** (0.0173)
Intercept	2.374** (0.0465)	2.439** (0.0484)	2.391** (0.0466)	2.4364** (0.0484)
Industry dummies	Included	Included	Included	Included
State dummies	Included	Included	Included	Included
Observations	216,768	216,768	216,768	216,768
R ²	0.566	0.566	0.566	0.566
Chi-square	545615.95	545554.65	545605.25	545544.95
Difference in chi-square relative to Model 2		−61.304**	−10.705**	−71.005**

Note. Robust standard errors clustered by state in parentheses; based on matching without replacement.

p < .10. *p < .05. **p < .01.

Robustness Tests

To test the robustness of Hypothesis 3, which is our main hypothesis, we first use the interaction of average wages (log of payroll per employee), rather than incentive compensation, with the family business variable. This is a reasonable test, since Chrisman et al. (2014) suggest that higher wages can also reduce the labor market sorting and attenuation disadvantage of family firms. As shown in Model 3, the interaction of the log of payroll per employee and family business variables is significant and positive (p < .01).

Second, we repeat this procedure, but this time we replace the incentive compensation variable with the benefits index. Although benefits in lieu of remuneration might not have as large of an impact as incentive compensation or higher wages, workers tend to understand that health insurance, retirement plans, and paid holidays are valuable substitutes for direct pay. Thus, as shown in Model 4, the interaction between the benefits index and the family business variable is significant and positive (p < .01).

Third, since incentive compensation, wages, and benefits may be seen as substitutes for one another, we include the interactions between the family business variable and the incentive compensation, payroll per employee, and benefits index variables in the same model. As seen in Model 5, the interactions between the (a) incentive compensation and family business variables and (b) payroll per employee and family business variables are both significant and positive (p < .01), as in Models 2 and 3, respectively. However, the interaction between the benefits index and the family business variable is not significant, supporting our conjecture that benefits are not as an effective remedy as incentive compensation (or higher base pay) for the labor market sorting and attenuation problem faced by family firms.

Fourth, we conducted a set of robustness tests based on the size of the firms. As with most U.S. Census data, fine-grained measures of firm characteristics are not available in the sample. We are therefore unable to assess the vertical distribution of incentive compensation (e.g., its distribution across lower-, mid-, and executive-level employees) or distribution of incentive compensation between family and nonfamily employees. However, as the size of a family firm increases, the proportion of nonfamily managers and other employees in a family firm should also increase, making the probability nonfamily employee participation in incentive compensation schemes greater. As presented in Table 6, based on the same specification in Tables 4 and 5, for firms with 20 or more employees and with 50 or more employees, the proposed relationships are significant (p < .01). While the inferences directly show that incentive compensation is positively associated with labor productivity in family firms as size increases, it indirectly suggests that similar to prior studies (e.g., Chrisman et al., 2007; Schulze et al., 2001) incentive compensation is unlikely to be confined to family members. Furthermore, as firm size increases the number of nonfamily managers will usually increase (Fang, Randolph, Memili, & Chrisman, 2016). Nevertheless, future research with multilevel data is necessary and inferences from Table 6 must be made with caution.

Table 6.

2SLS Regression Estimates by Size of Firm (Employees).

	(Model 6)	(Model 7)
	lnprod	lnprod
Family business	−0.1642** (0.0185)	−0.1600** (0.0183)
Incentive compensation	0.0668** (0.0089)	0.0680** (0.0090)
Incentive compensation × Family business	−0.0204 (0.0130)	−0.0028 (0.0112)
log of total employment	−0.1161** (0.0039)	−0.1510** (0.0042)
Dummy for ≥20 employees	0.0087 (0.0060)
Dummy for ≥20 employees × Incentive Comp. × Family business	0.0727** (0.0150)
Dummy for ≥50 employees		0.1516** (0.0096)
Dummy for ≥50 employees × Incentive Comp. × Family business		0.0682** (0.0134)
Log of payroll per employment	0.8046** (0.0056)	0.8045** (0.0056)
Log of payroll per employment × Family business	0.0367** (0.0059)	0.0352** (0.0058)
Benefits index	0.0266** (0.0034)	0.0317** (0.0035)
Benefits index × Family business	0.0055 (0.0044)	0.0051 (0.0044)
Seasonal business	−0.0056 (0.0172)	−0.0058 (0.0173)
Franchise	0.1035** (0.0101)	0.1018** (0.0102)
Homebased	−0.1071** (0.0090)	−0.1246** (0.0089)
Established between 1980 and 1989	−0.0198** (0.0056)	−0.0163** (0.0056)
Established between 1990 and 1999	0.0019 (0.0056)	0.0065 (0.0056)
Established between 2000 and 2002	0.0376** (0.0092)	0.0410** (0.0090)
Established in 2003	0.0054 (0.0109)	0.0076 (0.0109)
Established in 2004	0.0273** (0.0099)	0.0281** (0.0099)
Established in 2005	0.0237* (0.0113)	0.0244* (0.0111)
Established in 2006	−0.0578** (0.0110)	−0.0613** (0.0109)
Established in 2007	−0.1141** (0.0173)	−0.1319** (0.0171)
Intercept	2.4455** (0.0486)	2.5137** (0.0497)
Industry dummies	Included	Included
Industry dummies	Included	Included
Observations	216,768	216,768
R ²	0.566	0.567

Note. Robust standard errors clustered by state in parentheses; based on matching without replacement.

p < .10. *p < .05. **p < .01.

Finally, we test different matching procedures and found they yielded similar results. We further used a two-sample Kolmogorov–Smirnov test to determine whether there is a difference in the distribution of labor productivity for (a) firms with and without incentive compensation, (b) family and nonfamily firms, and (c) family and nonfamily firms with incentive compensation plans. The Kolmogorov–Smirnov tests show that in support of our hypotheses, the distributions are not equal (p = .000).

Discussion

We argue that family firms will have lower labor productivity than nonfamily firms because of adverse selection brought about by an attenuated labor market caused by workers’ sorting according to their self-interest. These conditions are a consequence of lower compensation and career opportunities provided by family firms, often because of the pursuit of noneconomic goals and idiosyncratic strategies by family owner-managers (Chrisman et al., 2014). Thus, we propose that the same factors that increase the risk of conflicts of interest between nonfamily managers and family owner-managers (Chua et al., 2009), may also increase the risk of adverse selection by reducing the attractiveness of working in a family firm for individuals who are not part of the owning family. However, we also suggest that the gap is narrowed when family firms use incentive compensation programs. Incentive compensation helps to align the interests of employees and owners, but this should be the case in both family and nonfamily firms alike. Therefore, while alignment of interests can help explain why incentive compensation increases labor productivity in family firms, it does not provide a sufficient explanation for why it increases labor productivity more in family firms than nonfamily firms. Indeed, as traditional agency theory suggests, the moral hazard problem emanating from conflicts of interest may be a less serious problem in family firms than in nonfamily firms (Chrisman et al., 2004; Fama & Jensen, 1983). Instead, we speculate that the greater impact of incentive compensation on family firms is due to a signaling effect that increases the likelihood that family firms will be able to attract high-quality workers and thereby diminish problems associated with adverse selection due to labor market sorting and attenuation.

We test our theory on a large sample of firms drawn from a survey conducted by the U.S. Census Bureau. The results show that our incentive compensation variable is positively related to higher labor productivity and family businesses have lower labor productivity than nonfamily businesses. After matching firms on a variety of factors, we find that family firms offering incentive compensation plans still have lower labor productivity than nonfamily firms that also use incentive compensation. However, the differential is significantly lower than when incentive compensation is not used. This suggests that in family firms, incentive compensation may not only align the interests of employees with that of family owner-managers, but it may also reduce the adverse effects of labor market sorting and attenuation by signaling a credible commitment that employees will be rewarded based on performance rather than merely kinship. Furthermore, our robustness tests using average wages and benefits indicate there is more than one way family firms can attract high-quality workers. Interestingly, as evidenced by the beta coefficients, the impact of incentive compensation is greater than the impact of average wages per employee, which is, in turn, greater than the impact of benefits packages. This is consistent with the theory because incentive compensation is more difficult for low-quality workers than for high-quality workers to obtain, and wages are a clearer and more direct signal of a firm’s commitment to performance than benefit packages.

The findings have important implications for family business theory and practice. Whether family businesses have higher (Martikainen, Nikkinen, & Vähämaa, 2009) or lower labor productivity (Barth et al., 2005; Lauterbach & Vaninsky, 1999; Wall, 1998) is an area of debate that relates to the broader issues of professionalization and relative performance (Stewart & Hitt, 2012). The large representative sample of family and nonfamily businesses in the current study helps address the issue and our theory helps explain why we obtained the answer we did. Although the large sample in the current study could lower the standard errors, and therefore increase significance of effects, effect sizes are not significantly influenced by large samples. As reported in Table 3, the effect sizes in the results are meaningful for family firms. Our findings indicate that incentive compensation and labor productivity in family firms are areas that may require greater attention from family owners. Labor productivity as an outcome is important to family firms striving for efficiency and profitability, and incentive compensation is one way they can motivate workers to increase their labor productivity. In general, incentive compensation can act as a mechanism that not only elicits value-increasing behaviors from current employees, but may also increases the odds of attracting high-quality employees who are most able to increase firm value.

Our findings also contribute to the compensation literature, which has a long history and, as a practice, appears to have become increasingly popular in recent years (Kruse et al., 2012). While past work has proposed several motives (Lee, 2006), to our knowledge we are one of the first to assess incentive compensation from a family business perspective and to explain its relevance in addressing the adverse selection problem. Specifically, our study suggests that incentive compensation can assist in recruiting qualified employees by signaling the willingness of a firm to reward performance. More generally, our findings imply that researchers should pay more attention to the impact of human resource practices on recruiting because effective human resource practices can potentially influence both the ability and the effort of an organization’s workforce. As our study indicates, this appears to be especially true for family firms. Owing to the importance of family firms to the world economy, the implications of incentive compensation and human practices in general to family firms needs further study by both family business and human resource scholars.

Limitations and Future Research

Our findings must be interpreted in the light of the limitations of this study. First, the size of the database as well as our robustness and endogeneity tests increase the relevance and reliability of our findings. However, the cross-sectional nature of the data allows us to demonstrate only correlation rather than causation. Longitudinal studies are therefore needed. Furthermore, it should be noted that although the measure of labor productivity is widely used in the literature, due to limitations of the data, we are unable to use alternate productivity measures or analytical methods such as data envelopment methods or stochastic frontier analysis. As such, our inferences are limited to labor productivity and not a firm’s total productivity. The inferences must also be made with the caveat that the results relate only to employees’ influence on sales and not to other outcomes (e.g., process improvements by employees that may not be directly reflected in sales).

Second, we measure family business as a dummy variable based on ownership. Recent studies have found strong correlations among multiple operationalizations of the family business variable and advocate dichotomous measures (Gomez-Mejia et al., 2014). Still, our study would have been stronger had we been able to conduct robustness tests using continuous measures of family involvement and influence.

Third, we focus only on a comparison of family and nonfamily businesses in the United States. The impact of incentive compensation may be different in other environments. Moreover, we do not have the data necessary to analyze the cumulative effects of incentive compensation across levels or between family and nonfamily employees. In Table 6, we attempt to make indirect inferences on the plausibility that incentive compensation plans are extended to nonfamily employees based on an analysis of firm size. However, further studies are needed to understand the extent to which the influence of incentive compensation on labor productivity might vary across hierarchical levels in an organization and between family and nonfamily employees.

Fourth, although we theorize that incentive compensation reduces problems of adverse selection associated with labor marketing sorting and attenuation, we were unable to measure the severity of that problem directly. Likewise, the reasoning we use to suggest family firms have greater adverse selection problems such as noneconomic goals, favoritism, and idiosyncratic strategies could not be directly tested, and of course, our discussion was never meant to imply that such behaviors or the resulting problems exist uniformly among family firms. We also lacked the data to control for other potentially important variables such as the composition of the management team and professionalization. While this problem is shared by many studies using secondary data, future work that permits more complete and direct measurement of these and other important variables distinguishing family and nonfamily firms are needed. Nevertheless, we were able to provide evidence that incentive compensation (as well as higher pay and benefits packages) has a greater influence on the labor productivity of family firms than on the labor productivity of nonfamily firms. We believe our comparison of the labor productivity of family and nonfamily firms provides a realistic test of our theory since the matched sampling methodology assists us in ruling out alternative explanations for our findings.

Besides eliminating these limitations, future research should assess the relative impact on performance of other types of compensation (both pecuniary and nonpecuniary) and human resource practices in family and nonfamily firms as well as their potential spillovers on other exchanges between family owners and employees. Likewise, it would be interesting to determine if monitoring, which might act as a signal of professionalization, has a similar impact on recruitment and, ultimately, the relative labor productivity of family and nonfamily firms as found for incentive compensation in this study. More generally, since alignment of interests imply the importance of fit between the interests of potential employees and owners, further research should investigate how signaling can be used to attract the kinds of workers a (family) firm needs and wants. Human resource studies that contribute to a better understanding of how workers interpret the conscious and unconscious signals that an organization transmits regarding its goals and values would be useful and this applies to both family as well as nonfamily firms. For example, Hauswald et al. (2015) find that job seekers who value conservation (tradition, conformity, security) and self-transcendence (universalism, benevolence) are more attracted to family firms than job seekers who value openness to change (self-direction, stimulation, hedonism) and self-enhancement (achievement, power). But more research is needed to understand how the relationship between organizational signals and the self-sorting of potential employees that lead them to favor one type of organization over another actually works.

Given that there could be a relationship between a family firm’s capacity to solve the adverse selection problem and its capacity to solve the alignment of interests problem, additional research on their overlap appears warranted. For example, although misrepresentation of ability could clearly contribute, adverse selection may also be based on problems of bounded rationality and information asymmetry caused by the self-sorting of self-interested potential employees. Thus, adverse selection can exist even in the absence of opportunism. Furthermore, since ability and effort are correlated, the effectiveness of a firm’s selection process could also plant the seeds for agency problems associated with moral hazard or the seeds of stewardship. This suggests that a firm’s ability to resolve the adverse selection problem caused by sorting and labor market attenuation may influence whether it develops a stewardship culture, a possibility stewardship theory is not currently able to deal with because it does not address either the adverse selection problem or the existence of bounded rationality and information asymmetry (Chrisman et al., 2014).

In closing, incentive compensation represents a costly but useful signal of family owners’ willingness to reward nonfamily employees for their contributions to the firm. As we have illustrated in this article, family firms realize a larger increase in labor productivity than nonfamily firms when they implement incentive compensation programs, which may come from the ability of incentive compensation to act as a signaling device that reduces adverse selection caused by sorting and attenuation in the labor market.

Footnotes

Acknowledgements

This article was awarded the 2016 Family Firm Institute Best Unpublished Paper Award. The authors thank Josh Daspit, Kincy Madison, Laura Marler, and James Vardaman for their comments on earlier versions of this article. The authors also thank the Associate Editor, Justin Webb, and two anonymous referees for their comments on the article during the review process.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

James J. Chrisman (PhD, University of Georgia) is the Julia Bennett Rouse professor of Management, head of the Department of Management and Information Systems, and director of the Center of Family Enterprise Research at Mississippi State University. He also holds a joint appointment as senior research fellow with the Centre for Entrepreneurship and Family Enterprise at the University of Alberta, School of Business.

Srikant Devaraj is a research assistant professor in the Center for Business and Economic Research at Miller College of Business, Ball State University. His research interests lie in the areas of applied econometrics, health economics, and health informatics. He earned a doctorate in economics from IUPUI.

Pankaj C. Patel is an associate professor of management at Villanova University. His research interests are at the intersection of family firm governance and innovation. He received his PhD from the University of Louisville.

References

Akerlof

G. A.

(1970). The market for “lemons”: Quality uncertainty and the market mechanism. Quarterly Journal of Economics, 84, 488-500.

Anderson

R. C.

Duru

Reeb

D. M.

(2012). Investment policy in family controlled firms. Journal of Banking & Finance, 36, 1744-1758.

Bach

Serrano-Velarde

(2015). CEO identity and labor contracts: Evidence from CEO transitions. Journal of Corporate Finance, 33, 227-242.

Bammens

Notelaers

Van Gils

(2015). Implications of family business employment for employees’ innovation work involvement. Family Business Review, 28, 123-144.

Bandura

(1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215.

Bandura

(1982). Self-efficacy mechanisms in human agency. American Psychologist, 37, 122-147.

Banks

(2008). Overview of risk management and alternative risk transfer. In Fabozzi

F. J.

(Ed.), Handbook of finance (Vol. III, pp. 39-52). Hoboken, NJ: Wiley.

Barbera

Moores

(2013). Firm ownership and productivity: A study of family and non-family SMEs. Small Business Economics, 40, 953-976.

Barnett

Kellermanns

F. W.

(2006). Are we family and are we treated as family? Nonfamily employees’ perceptions of justice in the family firm. Entrepreneurship Theory and Practice, 30, 837-854.

10.

Barth

Gulbrandsen

Schøne

(2005). Family ownership and productivity: The role of owner-management. Journal of Corporate Finance, 11, 107-127.

11.

Bennedsen

Pérez-González

Wolfenzon

(2010). The governance of family firms. In Baker

K. H.

Anderson

(Eds.), Corporate governance: A synthesis of theory, research, and practice (pp. 371-390). Hoboken, NJ: Wiley.

12.

Blinder

A. S.

(2011). Paying for productivity: A look at the evidence. Washington, DC: Brookings Institution Press.

13.

Block

J. H.

Fisch

C. O.

Lau

Obschonka

Presse

(2016). Who prefers working in family firms? An exploratory study of individuals’ organizational preferences across 40 countries. Journal of Family Business Strategy, 7(1), 65-74.

14.

Boling

Pieper

T. M.

Covin

(2016). CEO tenure and entrepreneurial orientation within family versus nonfamily firms. Entrepreneurship Theory and Practice, 40, 891-913.

15.

Caliendo

Kopeinig

(2008). Some practical guidance for the implementation of propensity score matching. Journal of Economic Surveys, 22(1), 31-72.

16.

Carney

(2005). Corporate governance and competitive advantage in family-controlled firms. Entrepreneurship Theory and Practice, 29, 249-265.

17.

Chandler

G. N.

McEvoy

G. M.

(2000). Human resource management, TQM, and firm performance in small and medium-size enterprises. Entrepreneurship Theory and Practice, 25, 43-57.

18.

Chrisman

J. J.

Chua

J. H.

Kellermanns

F. W.

Chang

E. P. C.

(2007). Are family managers agents or stewards? An exploratory study in privately held family firms. Journal of Business Research, 60, 1030-1038.

19.

Chrisman

J. J.

Chua

J. H.

Litz

R. A.

(2004). Comparing the agency costs of family and non-family firms: Conceptual issues and exploratory evidence. Entrepreneurship Theory and Practice, 28, 335-354.

20.

Chrisman

J. J.

Chua

J. H.

Steier

L. P.

Wright

McKee

D. N.

(2012). An agency theoretic analysis of value creation through management buy-outs of family firms. Journal of Family Business Strategy, 3, 197-206.

21.

Chrisman

J. J.

Memili

Misra

(2014). Nonfamily managers, family firms, and the winner’s curse: The influence of noneconomic goals and bounded rationality. Entrepreneurship Theory and Practice, 38, 1103-1127.

22.

Chua

J. H.

Chrisman

J. J.

Bergiel

E. B.

(2009). An agency theoretic analysis of the professionalized family firm. Entrepreneurship Theory and Practice, 33, 355-372.

23.

Chua

J. H.

Chrisman

J. J.

Sharma

(1999). Defining the family business by behavior. Entrepreneurship Theory and Practice, 23(4), 19-39.

24.

Cucculelli

Mannarino

Pupo

Ricotta

(2014). Owner-management, firm age, and productivity in Italian family firms. Journal of Small Business Management, 52, 325-343.

25.

Corbetta

Salvato

(2004). Self-serving or self-actualizing? Models of man and agency costs in different types of family firms: A commentary on “Comparing the agency costs of family and non-family firms: Conceptual issues and exploratory evidence.” Entrepreneurship Theory and Practice, 28, 355-362.

26.

Dehejia

R. H.

Wahba

(2002). Propensity score-matching methods for nonexperimental causal studies. Review of Economics and Statistics, 84, 151-161.

27.

De Kok

J. M.

Uhlaner

L. M.

Thurik

A. R

. (2006). Professional HRM practices in family owned-managed enterprises. Journal of Small Business Management, 44, 441-460.

28.

Dyer

W. G.

(2006). Examining the “family effect” on firm performance. Family Business Review, 19, 253-273.

29.

Eisenhardt

K. M.

(1989). Agency theory: An assessment and review. Academy of Management Review, 14, 57-74.

30.

Fama

E. F.

Jensen

M. C.

(1983). Separation of ownership and control. Journal of Law and Economics, 26, 301-325.

31.

Fang

Memili

Chrisman

J. J.

Penney

(2016). Industry and information asymmetry: The case of the employment of non-family managers in small and medium-sized family firms. Journal of Small Business Management. doi:10.1111/jsbm.12267

32.

Fang

Randolph

R. V. D. G.

Memili

Chrisman

J. J.

(2016). Does size matter? The moderating effects of firm size on the employment of nonfamily managers in privately held family SMEs. Entrepreneurship Theory and Practice, 40, 1017-1039.

33.

FitzRoy

F. R.

Kraft

(1987). Cooperation, productivity, and profit sharing. Quarterly Journal of Economics, 102(1), 23-35.

34.

Florkowski

G. W.

(1987). The organizational impact of profit sharing. Academy of Management Review, 12, 622-636.

35.

Gomez-Mejia

L. R.

Campbell

J. T.

Martin

Hoskisson

R. E.

Makri

Sirmon

D. G.

(2014). Socioemotional wealth as a mixed gamble: Revisiting family firm R&D investments with the behavioral agency model. Entrepreneurship Theory and Practice, 38, 1351-1374.

36.

Gomez-Mejia

L. R.

Haynes

K. T.

Núñez-Nickel

Jacobson

K. J.

Moyano-Fuentes

(2007). Socioemotional wealth and business risks in family-controlled firms: Evidence from Spanish olive oil mills. Administrative Science Quarterly, 52, 106-137.

37.

Hambrick

D. C.

D’Aveni

R. A.

(1988). Large corporate failures as downward spirals. Administrative Science Quarterly, 33, 1-23.

38.

Hancock

J. I.

Allen

D. G.

Bosco

F. A.

McDaniel

K. R.

Pierce

C. A.

(2013). Meta-analytic review of employee turnover as a predictor of firm performance. Journal of Management, 39, 573-603.

39.

Hart

R. A.

Huebler

(1991). Are profit shares and wages substitute or complementary forms of compensation? Kyklos, 44, 221-231.

40.

Hauswald

Hack

Kellermanns

F. W.

Patzelt

(2015). Attracting new talent to family firms: Who is attracted under what conditions? Entrepreneurship Theory and Practice, 40, 963-989.

41.

Jensen

M. C.

Meckling

W. H.

(1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305-360.

42.

Karra

Tracey

Phillips

(2006). Altruism and agency in the family firm: Exploring the role of family, kinship, and ethnicity. Entrepreneurship Theory and Practice, 30, 861-877.

43.

Kruse

D. L.

Blasi

J. R.

Freeman

R. B.

(2012). Does linking worker pay to firm performance help the best firms do even better? (Working Paper 17745). Cambridge, MA: National Bureau of Economic Research.

44.

Kruse

D. L.

Freeman

R. B.

Blasi

J. R.

(2010). Shared capitalism at work: Employee ownership, profit and gain sharing, and broad-based stock options. Chicago, IL: University of Chicago Press.

45.

Lauterbach

Vaninsky

(1999). Ownership structure and firm performance: Evidence from Israel. Journal of Management & Governance, 3, 189-201.

46.

Lee

(2006). Family firm performance: Further evidence. Family Business Review, 19, 103-114.

47.

Long

Fang

(2015). Do strategic factors affect adoption of profit sharing? Longitudinal evidence from Canada. International Journal of Human Resource Management, 26, 971-1001.

48.

Long

R. J.

(2000). Employee profit sharing: Consequences and moderators. Relations Industrielles/Industrial Relations, 55, 477-504.

49.

Martikainen

Nikkinen

Vähämaa

(2009). Production functions and productivity of family firms: Evidence from the S&P 500. Quarterly Review of Economics and Finance, 49, 295-307.

50.

Meade

J. E.

(1972). The theory of labour-managed firms and of profit sharing. Economic Journal, 82, 402-428.

51.

Memili

Misra

Chang

E. P.

Chrisman

J. J.

(2013). The propensity to use incentive compensation for non-family managers in SME family firms. Journal of Family Business Management, 3, 62-80.

52.

Michiels

Voordeckers

Lybaert

Steijvers

(2013). CEO compensation in private family firms: Pay-for-performance and the moderating role of ownership and management. Family Business Review, 26, 140-160.

53.

Miller

Le Breton-Miller

Scholnick

(2008). Stewardship vs. stagnation: An empirical comparison of small family and non-family businesses. Journal of Management Studies, 45, 51-78.

54.

Moore

W. J.

Newman

R. J.

(1985). The effects of right-to-work laws: A review of the literature. Industrial and Labor Relations Review, 38, 571-585.

55.

Morikawa

(2013). Productivity and survival of family firms in Japan. Journal of Economics and Business, 70, 111-125.

56.

Peikes

D. N.

Moreno

Orzol

S. M.

(2008). Propensity score matching. American Statistician, 62, 222-231.

57.

Rubin

D. B.

(2006). Matched sampling for causal effects. Cambridge, England: Cambridge University Press.

58.

Schulze

W. S.

Lubatkin

M. H.

Dino

R. N.

Buchholtz

A. K.

(2001). Agency relationships in family firms: Theory and evidence. Organization Science, 12, 99-116.

59.

Singh

(2004). Impact of HR practices on perceived firm performance in India. Asia Pacific Journal of Human Resources, 42, 301-317.

60.

Spence

(1973). Job market signaling. Quarterly Journal of Economics, 87, 355-374.

61.

Steinherr

(1977). On the efficiency of profit sharing and labor participation in management. Bell Journal of Economics, 8, 545-555.

62.

Stevans

L. K.

(2009). The effect of endogenous right-to-work laws on business and economic conditions in the United States: A multivariate approach. Review of Law & Economics, 5, 595-614.

63.

Stewart

Hitt

M. A.

(2012). Why can’t a family business be more like a nonfamily business? Modes of professionalization in family firms. Family Business Review, 25, 58-86.

64.

Verbeke

Kano

(2012). The transaction cost economics theory of the family firm: Family-based human asset specificity and the bifurcation bias. Entrepreneurship Theory and Practice, 36, 1183-1205.

65.

Wall

R. A.

(1998). An empirical investigation of the production function of the family firm. Journal of Small Business Management, 36, 24-32.

66.

Way

S. A.

(2002). High performance work systems and intermediate indicators of firm performance within the US small business sector. Journal of Management, 28, 765-785.

67.

Zavodny

(2000). The effect of the minimum wage on employment and hours. Labour Economics, 7, 729-750.