1042-2587 © 2008 by Baylor University

E T&P

The Effect of Small Business Managers’ Growth Motivation on Firm Growth: A Longitudinal Study
Frédéric Delmar1 Johan Wiklund

This study addresses the role of small business managers’ growth motivation for business growth, taking into account the important effects of previous motives and feedback from earlier performance. We hypothesize that small business managers’ growth motivation has a unique influence on firm outcome measured as growth in sales and in number of employees. Data were gathered from two different Swedish samples of small firms using telephone interviews. Using cross-lagged regression analysis, we find support for our hypotheses when examining employment growth, but only partial support when examining sales.

Introduction
The psychological construct of motivation has an important role to play in entrepreneurship research. As stated by Shane, Locke, and Collins (2003, p. 257): “We believe that the development of entrepreneurship theory requires consideration of the motivations of people making entrepreneurial decisions.” One of the areas in entrepreneurship where motivation is potentially of great importance relates to firm growth. There is research to suggest that growth is one of the most important outcomes of entrepreneurial efforts because it indicates the degree of success of that effort (Bhidé, 1999; Venkataraman, 1997), and effort exerted is closely related to the individual’s motivation (Davidsson, Delmar, & Wiklund, 2002). Research examining the link between growth motivation and growth appears to support this view as it finds a positive relationship between growth motivation and growth (e.g., Baum, Locke, & Kirkpatrick, 1998; Baum, Locke, & Smith, 2001; Kolvereid & Bullvag, 1996; Miner, Smith, & Bracker, 1989). Implicitly, the view underlying this research and the theories used is the assumption that growth motivation affects the future growth of the firm, i.e., that growth motivation has a causal effect on firm growth. However, in their review and test of leading theories on
Please send correspondence to Johan Wiklund, tel.: +1 315-443-3356; fax: +1 315-442-1449; e-mail: jwiklund@syr.edu 1. Both authors contributed equally and are listed alphabetically.

May, 2008

437

goal-directed behaviors, Bagozzi and Kimmel (1995) demonstrated that these theories are incomplete because they fail to consider the feedback from past behavior and behavioral outcomes. Drawing on these findings, later research has further elaborated on the relationship between motivation and behavior (see e.g., Conner, Sheeran, Norman, & Armitage, 2000; Ouellette & Wood, 1998; Perugini & Conner, 2000; Sheeran & Abraham, 2003; Triandis, 1977). Specifically, the temporal stability of motives has been investigated, with results indicating that stable motives are good predictors of behavior (Sheeran & Abraham, 2003). These recent theoretical developments open up for the possibility that both motivation and future behavior represent reactions to past behavior and outcomes rather than being the result of the commitment to specific motives (e.g., Ouellette & Wood, 1998), thus challenging the causal structure of motivational models in entrepreneurship research. To our knowledge, no research has considered how the outcomes of past behavior and the stability of motives affect the relationships between the motivations of small business managers and future outcomes. Such research is important because the failure to recognize the influence of past behavior and temporal stability of motives could lead to misinterpretations of the causal effects of motivations on outcomes. This has implications for how we model the relationship between motivation and outcomes in entrepreneurship research. Explicating and testing the causal structure of research models constitute an important step in theory development (see e.g., Whetten, 1989 for a discussion of theoretical contributions and Gartner, 1989 for a specific discussion on entrepreneurship theory and theory development). Therefore, the purpose of this paper is to examine the causal direction between growth motivation and firm growth. We test to what extent the motivation of small business managers affect growth accounting for the feedback from previous growth using longitudinal data and cross-lagged regression analyses. The paper continues as follows. Next we provide a short summary of the theoretical rationale for the study. We then review relevant motivation–outcome research and develop our hypotheses. This is followed by a section on methodology, which describes our sample and variables and analyses and results. Finally, we discuss the findings and draw the theoretical and empirical implications of the research.

Theory and Hypotheses Temporal Stability of Growth Motivation
Motivation theories build on the premise that our motivations affect our behavior. Motivation affects the choice of behavior, the longevity of the behavior, and the level of effort (Kanfer, 1991). The growth motivation of a small business manager, defined as the aspiration to expand the business, reflecting attitudes and subjective norms in Ajzen’s (1991) theory of planned behavior, affects his or her choice to expand the business, the willingness to sustain this choice over time, and at what level of effort. As opposed to many other behaviors and outcomes studied in the motivation literature, growth is not instantaneous but a process that unfolds over time (Penrose, 1959). Empirical studies of firm growth have typically assessed periods varying from 1 or 2 years up to 5 years (Wiklund & Shepherd, 2005). Over such relatively long time periods, there are multiple activities, actions, and decisions that affect the firm growth process. If the small business manager is motivated to expand his or her firm during a short period of time only but later prioritizes other goals and behaviors—for example because the efforts put into expanding the business did not lead to the desired outcomes (McCloy, Campbell, & Cudeck, 1994)—
438 ENTREPRENEURSHIP THEORY and PRACTICE

there is likely little effect of growth motivation on actual growth during extended periods of time. Unless motivation remains relatively constant over time until the behavior is performed, prediction will be weak (Ajzen, 1995). Empirically, research has found that stable motivations are good predictors of behavior while unstable motivations have no association with behavior (Sheeran, Orbell, & Trafimow, 1999). Therefore, an implicit assumption in the literature on growth motivation is that this motivation remains relatively stable over time. The stability of growth motivation is central to theoretical development and to our subsequent hypotheses because the argument for the relationship between motivation and growth hinges on relative stability of growth motivation, but it has not been empirically assessed in the literature. Some indirect indications of stability may be inferred from previous research. A few studies have assessed the relationship between growth motivation and growth using a time lag between motivation and growth and have found support for a positive association (e.g., Baum et al., 2001; Kolvereid & Bullvag, 1996). Unless these findings are spurious, it suggests that growth motivation has some stability over time. This allows us to formulate our first hypothesis: Hypothesis 1: Growth motivation at time1 (T1) has a positive effect on growth motivation at time2 (T2).

The Effect of Growth Motivation on Growth
The effect of motivation on behavior is less obvious than may be assumed for two primary reasons. First, the strength of the relationship is affected by the individual’s degree of volitional control, i.e., the ability to perform the behavior at will. Environmental constraints and insufficient ability or task comprehension (i.e., not understanding what to do) diminish the effect of motivation on behavior. For example, the greater an individual’s ability, the greater is his or her tendency to choose to act (Kanfer, 1991; McCloy et al., 1994). Limited volitional control has been incorporated into goal-directed behavioral theories. The theory of planned behavior is an extension of the theory of reasoned action (Ajzen & Fishbein, 1977), adding aspects of individual ability (Ajzen, 1991), thus incorporating behaviors over which people have incomplete volitional control. As expected, this theory has been shown to outperform the theory of reasoned action in situations of limited volitional control (Netemeyer, Burton, & Johnston, 1991). The expansion of a firm is an example of a behavior that is under limited volitional control. Unless the managers of the firm have the ability to develop suitable strategies and can spot growth opportunities, the firm will not grow irrespectively of the motivation to expand the business (Wiklund & Shepherd, 2003). Second, the complexity of the behavior also affects its relationship with motivation. The expansion of a business is complex and can therefore be considered a fuzzy task (Campbell, 1988), characterized by multiple ways of attaining the desired outcome. It is also characterized by uncertainty and interdependences. This means that small business managers can choose many different ways to achieve growth and the goal of growing the firm can be interdependent with other goals. For example, small business managers may want to expand their firms but can achieve this desired outcome in several different ways; they can acquire another firm or grow organically; grow by increasing sales while outsourcing the production, thus expanding sales but not employment. Furthermore, if the goal of expanding the business conflicts with other goals, the small business manager may choose actions that do not completely fulfill the expansion goal. Therefore, motivation theories suggest that growth motivation should have a positive effect on growth, but the effect could not be expected to be very large given that the
May, 2008 439

behavior is under limited volitional control and that the task of expanding a business could be regarded as complex and fuzzy. Previous empirical research tends to confirm this (e.g., Baum et al., 2001; Bellu & Sherman, 1995; Kolvereid & Bullvag, 1996; Mok & Van den Tillaart, 1990). These studies find some support for a positive relationship between motivation and growth, although these relationships are generally not very strong compared to what has been found for less complex behaviors under greater volitional control (see, e.g., Armitage & Conner, 2001 for a meta-analysis of the relationship between motivation and behavior in the theory of planned behavior). This leads us to anticipate a positive (but relatively weak) relationship between growth motivation and growth. Thus: Hypothesis 2: Growth motivation at T1 has a positive effect on growth at T2.

The Feedback of Past Growth on Growth Motivation
Bagozzi and Kimmel (1995) suggest that a shortcoming of theories of goal-directed behavior is their failure to properly consider the role of past behavior. The possible feedback of previous behaviors and outcomes on future motivation is an important aspect of motivation theories because it has consequences for the modeling of the causal order of the two constructs. These authors suggest that past behavior affects the motivation to perform the behavior in the future. While it is true that past behavior is not emphasized by the theories, feedback from outcomes to motivation have been recognized by the theory of planned behavior (Ajzen, 1991), Bandura’s (1986, 1991), social cognitive theory as well as attribution theory (Anderson, 1991; Thomas & Mathieu, 1994; Weiner, 1985). In general, if previous behavior is perceived as successful by an individual, we would expect that the future motivation of that individual to perform the task increases. On the other hand, if the previous behavior is perceived as a failure, we would expect the motivation to decrease for the task or to become directed toward another target (cf. Bagozzi & Warshaw, 1992). In the context of this paper, this means that a small business manager who has experienced firm growth and attributes that outcome to his or her own ability will get a higher growth motivation than another small business manager who has not experienced firm growth or does not attribute that outcome to his or her own ability. The actual growth outcome is an important indicator of the entrepreneur’s ability to manage and expand the firm. If the outcome is positive, motivation will be reinforced. If the outcome is negative, the entrepreneur’s motivation will be reduced (Orpen, 1980; Silver, Mitchell, & Gist, 1995; Thomas & Mathieu, 1994).2 Thus, these theories suggest that to some extent, growth motivation can be seen as an “acquired taste” (cf. Davidsson et al., 2002) resulting from past growth outcomes. This argument leads to the following hypothesis: Hypothesis 3: Growth at T1 has a positive effect on growth motivation at T2.

The Independent Impact of Motivation
As illustrated in previous discussions, it is fair to say that several different claims about the association between growth and growth motivation have been made in the literature. Calls for research that makes a more detailed examination of the interplay
2. The escalation of commitment literature suggests that the opposite is also possible, i.e., that people may be enticed to commit even further resources to a failed course of action (e.g., Brockner, 1992). However, we follow the major line of reasoning in the motivation literature in stating our hypothesis.

440

ENTREPRENEURSHIP THEORY and PRACTICE

Figure 1 Cross-Lagged Analysis of Growth and Growth Motivation
Time1
Growth motivationT1

Time2 a b c
Growth motivationT1

GrowthT1

GrowthT2

d a > 0 = hypothesis 1 b > 0 = hypothesis 2 c > 0 = hypothesis 3 b > d = hypothesis 4 between growth and growth motivation have been made in the literature (Wiklund & Shepherd, 2003). Bagozzi and Kimmel (1995) noted that empirical applications of motivation theories often failed to take into account the effect of past behavior, which had profound effects on the relationship between motivation and behavior. Specifically, they showed that past behavior in some models was a much better predictor of future behavior than was motivation and that the inclusion of past behavior even potentially reduced the effect of motivation to zero. If these results hold up in the present context, past growth would be a better predictor of future growth than would growth motivation. If so, there is little reason to rely on motivation theories if the primary interest is to investigate factors that influence the growth of small firms. Moreover, the application of motivation theories to the context of small firm growth would seem superfluous. We, therefore, test which has a greater influence on growth—past growth or growth motivation. Motivation theories hold that the intentional behavior of individuals has a strong independent impact on the future development of their firms (Ajzen, 1995; Bandura, 1982; Weiner, 1992). Siding with these theories, we hypothesize that even if we include past growth in the model, the effect of growth motivation is stronger than the effect of past growth. Thus: Hypothesis 4: Growth motivation at T1 has a stronger impact on growth at T2 than has growth at T1. The relationships that we test and our hypotheses are summarized in Figure 1.

Method Research Design and Samples
Two samples were combined to test the hypotheses. Both sample frames were taken from Statistics Sweden’s data on incorporated companies in Sweden. By law, all incorporated companies have to report the data of this database. The samples were stratified
May, 2008 441

Table 1 Development of the Two Samples across the Period of Observation
First round
Sample 1 Data collected year Number sampled Number interviewed Number of known exits Sample 2 Data collected year Number sampled Number interviewed Number of know exits Combined Total number of cases

Second round

1994 730 400 (54.8%)

1998 400 314 (78.5%) 29 1999 630 549 (87.1%) 52 863

1996 808 630 (78.0%)

1030

over the Swedish equivalent of ISIC codes (high-technology manufacturing, lowtechnology manufacturing, services, and professional services), and standard Swedish size brackets: 5–9, 10–19, and 20–49 employees in the first study and 10–19 and 20–49 employees in the second study. The samples are detailed in Table 1. The data were collected over the telephone from the managing director, who in most cases was also the majority owner. The managing director was explicitly asked for at the beginning of the interview. The exact same questions were asked in all interviews. In both samples, data were collected twice, measuring growth and growth motivation at both points in time. In a few cases, the managing director had been replaced during the time between the two interviews. These cases were excluded from the analyses. In the first study, data were collected in 1994 and 1998; in the second study they were collected during 1996 and 1999. Thus, the first study had a time span of 4 years between the waves and the second study had a 3-year time span. Response rates were notably high (54.8– 87.1%), which helps safeguard against nonresponse bias. The collection of data at two points in time reduces the problem of memory decay and cognitive bias as well as common method variance. It also allows for testing of causal order between the constructs using cross-lagged regressions. Granger causality holds that an effect in a prior period can lead to an effect in a subsequent period; however, an effect in a subsequent period cannot lead to an effect in a prior period. As a result of the different size brackets included in the two studies, there are some significant differences between the two samples. Sample 1 includes firms that are significantly smaller (mean = 18.1 employees, standard deviation [SD] = 39.4 vs. mean = 23.0 employees, SD = 14.1) and younger (mean = 12.1 years, SD = 10.1 vs. mean = 29.4 years, SD = 27.7) than sample 2.

Measures
Previous research suggests that growth in terms of employment and sales are important growth indicators that provide different and complementary information (Delmar,
442 ENTREPRENEURSHIP THEORY and PRACTICE

1997; Delmar, Davidsson, & Gartner, 2003; Weinzimmer, Nystrom, & Freeman, 1998). Therefore, we relied on both these aspects concerning growth and growth motivation. Employment and Sales Growth. During all interviews we asked for present size in terms of employees (converted into full-time equivalents) and annual sales. We also asked for the corresponding figures 3 years ago. We explicitly tested these recall questions for potential recall bias. In the 1999 round of the second study, these “3 years ago” questions correspond to the “present size” figures reported during the first round in 1996. We correlated these figures. The correlations of both sales and employment were above .95, which ensures that respondents were able to correctly recall the size of their firms 3 years ago. GrowthT1 was calculated as the relative size change between the present size reported during the first round and the size 3 years prior. GrowthT2 was calculated as the relative size change between the two rounds. Data were heavily skewed, containing several outliers. Several techniques exist to normalize data (e.g., logarithmic transformation). We relied on the Winsor technique (e.g., Kennedy, Lakonishok, & Shaw, 1992), where a fixed percentage (in our case 5%) of the outlier cases at the tale of the distribution receive the same values as the observations at the truncation point (i.e., the 95 percentile). This transformation led to a distribution that was acceptable, with a minimum change in the data. Sales and Employment Growth Motivation. At each interview, we asked the respondents: “If the firm develops the way you would like it to, how many employees and how many large sales would the firm have 5 years ahead? Disregard possible inflation.” Based on these responses, growth motivation was calculated as the relative difference between intended size and current size in terms of employment and sales. Growth MotivationT1 was tapped during the first interview rounds and Growth MotivationT2 was tapped during the second. Again, Winsorization was used to normalize the data. This measure is similar to what has been used in previous studies of growth motivation. Kolvereid and Bullvag (1996) refer to a similar variable as growth intention; Wiklund and Shepherd (2003) call it growth aspiration and Wiklund, Davidsson, and Delmar (2003) call it attitude toward growth. All these authors relate to Ajzen’s (1995) theory of planned behavior. We argue that calling it an intention is overreaching, as the measure has no component of intended effort. Rather, we suggest that the concept represents a growth aspiration, which reflects attitudes and subjective norms in Ajzen’s theory. Two other possible measures of growth motivation were also available: (1) whether a 25% increase in the number of employees in 5 years’ time would be mainly negative or mainly positive and (2) whether a 100% increase in the number of employees in 5 years’ time would be mainly negative or mainly positive. While it would have been possible to utilize either of these alternative measures of the dependent variable or to compute a global growth motivation index, we prefer to rely on the question concerned with growth motivation in terms of growth rates because it makes the measurement scales for growth motivation and growth symmetrical (relative growth rates). This sort of symmetry in independent and dependent variables is deemed important by motivation theories (e.g., Ajzen & Fishbein, 1980; Eagly & Chaiken, 1993). In order to validate the measure we created a global growth intention index consisting of our two growth motivation variables and these two items from the 25% and the 100% scales. The Cronbach’s alpha value of the index was .72 and corrected item-total correlations ranged from .47 to .55, indicating that the index has acceptable reliability (Nunnally, 1967) and that all items share sufficient variance with the index (Nunnally & Bernstein, 1994). This index was also successfully used in predicting actual
May, 2008 443

growth outcomes in another study (Wiklund & Shepherd, 2003). This suggests that our growth motivation variables measure is (1) sufficiently reliable and (2) predictively valid (Nunnally & Bernstein, 1994). We therefore feel confident in relying on the chosen measures for growth motivation.

Control Variables
Previous research suggests that the individual characteristics of the small business manager; firm size and age, and industry affects growth (Davidsson, 1989; Delmar, 1996; Wiklund, 1998). We asked the respondent to state his or her highest completed education and constructed dummy variables for elementary school, trade school, high school, and university degree (university degree being the base category), which covers the vast majority of respondents. We also asked how the respondent became CEO of the firm and created dummy variables for the categories started, bought, inherited, or other. We asked what year the respondent was born, which was recoded into age at the time of the interviews. Based on the sampling frame, dummy variables were constructed for the four industries sampled (high-technology manufacturing, low-technology manufacturing, services, and professional services). The number of employees during the first survey round was included to measure size. Finally, we asked if the respondent knew which year the firm was founded. Their responses were recoded into firm age at the time of the interviews.

Correcting for Selection Bias
Several of the small firms failed between the first and second survey rounds and others did not complete all survey rounds. Because the variables that have an effect on growth may also impact business failure, attrition may not be random but instead systematic in a fashion similar to self-selection (Heckman, 1979). Therefore, unless the results are corrected for attrition, results may have a survivor bias. We therefore used the Heckman-type correction models (cf. Heckman, 1979; Shaver, 1998). Heckman-type correction models use a two-step process. First, based on theory, a model is developed for the probability of survival, which can predict this probability for each case. To develop the model for survival (selection model), we used research findings on factors contributing to small business survival and high performance (Wiklund, 1998). Specifically, firm age, firm size, and industry were used in the selection model (these variables also appear in the models for predicting the dependent variables). Second, a correction is made for self-selection (survival) by incorporating these predicted individual probabilities into the estimation model. A significant Rho suggests that if a correction had not been made then self-selection would have likely biased the results.

Analysis
First, the study’s measures were investigated to assess the quality of the data. Second, we compared the Pearson correlations between growth motivation and growth with previous research. Third, the hypothesized relationships were investigated using crosslagged regressions to embed the variables within the proposed research model as a means of gaining further insights regarding causal relationships. Our hypotheses suggest relatively complex causal relationships between growth motivation and growth. Specifically, they state dual causality between growth motivation
444 ENTREPRENEURSHIP THEORY and PRACTICE

and growth, suggesting that motivation affects growth and that there is feedback from achieved growth on future growth motivation. In order to test these hypotheses, we relied on cross-lagged regression analysis. Cross-lagged regressions are used to determine the causal relationship between constructs that are measured at least twice (Cohen & Cohen, 1983). It has been used in several studies to untangle the causal relationship between constructs where correlations have been noted in past research. For example, studies have examined the causal relationship between employee satisfaction and organizational effectiveness (Koys, 2001), between communication effectiveness and innovativeness (Lind & Zmud, 1991), or if political talk shows influence the audience or whether the audience is merely selecting sources consistent with previous political attitudes (Yanovitzky & Capella, 2001). The cross-lagged model we used is displayed in Figure 1 together with the hypotheses. As illustrated in the figure, two separate analyses were performed. First, growth at T2 (second survey) was regressed on growth at T1 and growth motivation at T1 (first survey). Second, growth motivation at T2 (second survey) was regressed on growth at T1 and growth motivation at T1 (first survey). In addition, a comparison of the size of the regression coefficients tells us which causal relationships were stronger. Since we measured growth and growth motivation in both employment and sales, this gave us a total of four different regression analyses.

Results
Table 2 reports means, SDs, and correlations between all variables. The low correlations provided initial evidence of discriminant validity. The correlation between growth motivation and actual growth was .27 for employment and .29 for sales. As a first assessment of our results, we compared these values to previous empirical studies investigating the relationship between growth motivation and growth using a longitudinal design. Mok and Van den Tillaart (1990) reported cross-tabulations of growth expectations and relative sales growth. Converting this to a correlation using the formula suggested by Rosenthal (1991) it became .42. Miner et al. (1994) reported a correlation of .48 for task motivation and absolute sales growth, and a correlation of .47 for task motivation and absolute employment growth. Bellu and Sherman (1995) found a correlation of .43 between task motivation and relative sales growth. Kolvereid and Bullvag (1996) reported mean differences between growth-oriented and not growth-oriented entrepreneurs. Converting the means to correlations (Rosenthal, 1991), they became .12 for sales and .16 for employment. Thus, our relationships between growth motivation and growth appeared to be of a similar magnitude compared to what has been found in previous research.

Predicting Growth Motivation
Tables 3 and 4 present the results from the hierarchical cross-lagged regressions performed for growth motivation and firm growth, respectively. The last rows in the models present the correlation of the error terms of the selection model and the prediction model (Rho). Then follows model fit (c2 and log likelihood), the increase thanks to the addition of the research variables (Dc2), and the number of cases included. Due to internal missing values the number of cases varied across the models. It could be noted in Table 3 that Rho was significant in the prediction of growth motivation, suggesting that results would have been biased had we not corrected for sample selection.
May, 2008 445

446

Table 2

Descriptive Statistics
2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12) 13) 14) 15) 16) 17) 18) 19) 20) 21) 22)

Variable

Mean SD

1)

1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12) 13) 14) 15) 16) 17) 18) 19) 20) 21) 22) 1.00 .13 .75 .10 .31 .09 .29 -.11 -.10 -.06 .01 -.05 .01 .03 -.05 .04 -.04 -.03 .00 -.03 .09 1.00 .09 .09 .15 .09 .15 -.18 -.09 .00 .05 -.15 -.06 -.06 -.09 .03 -.03 .00 -.01 -.19 .06 1.00 .17 .32 .16 .39 -.11 -.11 -.06 .01 -.02 .04 .00 -.02 .00 .02 -.01 -.05 .00 .10 1.00 .07 .76 .10 -.18 -.01 .08 .00 -.21 .27 -.10 -.15 .02 .04 .07 -.16 .13 .05 1.00 .08 .65 -.15 -.07 -.02 .01 -.19 -.14 -.04 -.05 .01 -.07 .02 .01 -.18 .01 1.00 .12 -.18 -.02 .06 -.03 -.14 .26 -.09 -.10 .03 .05 .08 -.15 .14 .07 1.00 -.12 -.12 -.04 -.02 -.15 -.12 -.06 .00 -.04 .01 .04 -.02 -.21 .04 1.00 .22 -.01 -.09 .15 .00 .08 -.07 -.14 .05 .06 -.03 .08 -.11 1.00 -.15 -.35 .02 -.07 -.04 .00 -.13 .03 .08 -.01 .06 -.11 1.00 -.25 -.02 -.05 .00 .02 -.03 .08 .03 -.07 .12 -.07 1.00 -.05 -.02 -.01 .07 -.05 .05 .08 -.06 -.06 -.04 1.00 .12 .15 .30 .10 .03 .01 .03 .35 .01 1.00 .01 -.04 .16 -.02 .07 -.04 .36 .20 1.00 -.24 -.18 .01 .05 -.05 -.07 -.18 1.00 -.17 .07 .03 -.02 -.01 -.12

Emp growthT2 Emp growth motT2 Sales growthT2 Sales growth motT2 Emp growthT1 Emp growth motT1 Sales growthT1 Sales growth motT1 Age of CEO T1 Elementary school High school Trading school Age of firm T1 Number of emp T1 Firm bought Firm heritage Firm other High tech Low tech Prof serv Origin of sample Subsidiary T1

1.17 1.37 1.25 1.63 1.29 1.46 1.58 1.70 47.73 .17 .10 .36 25.50 19.89 .26 .16 .20 .28 .26 .24 1.72 .13

.44 .46 .48 .62 .51 .49 .67 .61 8.79 .38 .30 .48 25.58 11.37 .44 .36 .40 .45 .44 .43 .45 .34

1.00 .08 .72 .12 .07 .27 .12 .10 -.18 -.09 -.01 .09 -.15 -.10 -.07 -.08 .02 -.04 -.03 .01 -.13 .03

1.00 -.16 -.09 .07 .26 .50

1.00 -.37 -.35 -.01 -.13

1.00 -.33 .02 -.07

1.00 -.05 .16

1.00 .24

1.00

ENTREPRENEURSHIP THEORY and PRACTICE

Note: N = 673, Correlations greater than .09 are significant at p < .05; those greater than .11 are significant at p < .01; and those greater than .13 are significant at p < .001. SD, standard deviation.

Table 3 Regression Analysis with Heckman Selection for Sales Growth MotivationT2 and Employment Growth MotivationT2
Sales Growth MotivationT2 Model 1 B
Age of CEO T1 Elementary school High school Trading school Age of firm at T1 Number of emp. T1 Firm bought Firm heritage Firm other High tech Low tech Prof services Origin of sample Subsidiary at T1 GrowthT1 Growth motivationT1 Constant Model Rho log likelihood c2 Dc2 Uncensored obs. Censored obs. -.008 -.146 -.200 -.121 .000 .000 .017 -.011 -.164 .074 .089 -.088 .212 .195

Employment Growth MotivationT2 Model 3 B
-.005 -.103 -.125 -.065 .000 -.001 .058 -.004 -.046 -.015 .020 -.038 .139 .107

Model 2 B
-.004 -.092 -.142 -.044 .001 .000 .058 -.007 -.140 .028 -.006 -.075 .221 .141 .081 .301 .650 .85*** 981 124*** 158*** 698 176

Model 4 B
-.002 -.088 -.094 -.039 .001 -.001 .085 .015 -.043 -.019 -.027 -.033 .120 .094 .094 .249 .687 .84*** 836 97*** 118*** 727 176

(SE)
(.002)** (.064)* (.079)* (.049)* (.001) (.002) (.052) (.066) (.068)* (.074) (.074) (.074) (.065)** (.074)**

(SE)
(.003)† (.064) (.078)† (.050) (.001) (.002) (.052) (.066) (.067)* (.069) (.071) (.070) (.062)*** (.075)† (.033)* (.037)*** (.188)***

(SE)
(.002)* (.049)* (.060)* (.039)† (.001) (.002) (.040) (.052) (.053) (.055) (.055) (.055) (.048)** (.059)†

(SE)
(.002) (.048)† (.060) (.038) (.001) (.002) (.040)* (.051) (.052) (.053) (.053) (.053) (.047)* (.058) (.034)** (.033)*** (.145)***

1.526 .92*** 1,060 51*** 735 176

(.162)***

1.286 .87*** 895 31** 757 176

(.126)***

†

p < .10; * p < .05; ** p < .01; *** p < .001

Model 1 is the base model including the control variables predicting sales growth motivationT2. The model was statistically significant (c2 = 51; p < .001). The regression coefficients show that CEO age had a negative effect on growth motivation. The negative effect of all education dummy variables suggested that those with the longest education (university degree = base category) had the highest growth motivation. The “Firm other” category, i.e., those who neither started, inherited nor bought the firm (e.g., employed CEOs) showed the lowest growth motivation. Higher growth motivation was also noted for respondents in the second sample (origin of sample) and firms that were subsidiaries within business groups. The research variables were added in model 2. These additions improved significantly the model fit (Dc2 = 158; p < .001). The effect of growth motivationT1 on growth motivationT2 was positive and statistically significant (b = .301; p < .001), providing partial support for hypothesis 1. The effect of growthT1 on growth motivationT2 was also
May, 2008 447

Table 4 Regression Analysis with Heckman Selection for Sales GrowthT2 and Employment GrowthT2
Sales GrowthT2 Model 5 B
Age of CEO T1 Elementary school High school Trading school Age of firm at T1 Number of emp. T1 Firm bought Firm heritage Firm other High tech Low tech Prof services Origin of sample Subsidiary at T1 GrowthT1 Growth motivationT1 Constant Model Rho Log likelihood c2 Dc2 Uncensored obs. Censored obs. -.006 -.060 -.025 -.003 .000 .000 -.076 -.115 .025 -.002 .057 -.045 -.057 .066

Employment GrowthT2 Model 7 B
-.005 -.078 -.056 .039 -.001 -.003 -.071 -.096 -.013 -.008 .034 -.021 .075 .003

Model 6 B
-.005 -.024 .002 .032 .000 -.001 -.057 -.109 .031 -.018 .027 -.034 -.071 .057 .057 .072 1.310 .72 967 34*** 72*** 782 176

Model 8 B
-.004 -.053 -.037 .056 .000 -.002 -.071 -.088 -.018 -.001 .011 -.004 .047 .006 .051 .197 .922 .62 903 72*** 88*** 803 176

(SE)
(.002)** (.051) (.060) (.040) (.001) (.002) (.041)† (.054)* (.054) (.054) (.053) (.053) (.048) (.060)

(SE)
(.002)* (.053) (.062) (.042) (.001) (.002) (.043) (.055)* (.057) (.055) (.055) (.054) (.052) (.064) (.029)* (.030)* (.156)***

(SE)
(.002)** (.046)† (.055) (.037) (.001) (.002)† (.038)† (.049)† (.050) (.049) (.048) (.049) (.043)† (.056)

(SE)
(.002)* (.047) (.056) (.037) (.001) (.002) (.038)† (.050)† (.050) (.047) (.048) (.047) (.044) (.057) (.033) (.033)*** (.143)***

1.538 .78* 1,003 29** 824 176

(.132)***

1.309 .77** 947 37*** 838 176

(.119)

†

p < .10; * p < .05; ** p < .01; *** p < .001

positive and statistically significant (b = .081; p < .05), providing partial support for hypothesis 3. Model 3 is the model for employment growth motivation with only the control variables. The results of the base model were quite similar to those obtained for sales growth motivation. The base model including the control variables was statistically significant (c2 = 31; p < .01). The signs of the coefficients were typically the same. A negative statistically significant coefficient could be noted for CEO age. Again, those with the longest education (university degree) had the highest growth motivation. Higher growth motivation was again noted for respondents in the second sample (origin of sample). In model 4 we entered the research variables. The addition gave a statistically significant improvement to the model fit (Dc2 = 118; p < .001). The effect of growth motivationT1 on growth motivationT2 was positive and statistically significant (b = .249;
448 ENTREPRENEURSHIP THEORY and PRACTICE

p < .001), providing additional support for hypothesis 1.3 The effect of growthT1 on growth motivationT2 was also positive and statistically significant (b = .094; p < .01), providing additional support for hypothesis 3. Thus, hypotheses 1 and 3 were supported by both analyses.

Predicting Firm Growth
We now turn to the analysis of firm growth. The results are shown in Table 4. Model 5 was statistically significant (c2 = 29; p < .01). It is the base model with the control variables included that predict sales growth. The regression coefficients show that CEO age had a negative effect on sales growth. Similarly, a negative effect was noted for those who have inherited their firm. In model 6, we added the research variables to the model and we noted a statistically significant improvement in model fit (Dc2 = 72; p < .001). The effect of growth motivationT1 on growthT2 was positive and statistically significant (b = .072; p < .05), providing partial support for hypothesis 2. The effect of growthT1 on growthT2 was also positive and statistically significant (b = .057; p < .05). The results of model 7 for employment growth were quite similar to those obtained for sales growth. This base model including the control variables was statistically significant (c2 = 37; p < .001). The only statistically significant coefficient was the negative effect of CEO age. Adding the research variables in model 8 made a statistically significant improvement to model fit (Dc2 = 88; p < .001). The effect of growth motivationT1 on growthT2 was positive and statistically significant (b = .197; p < .001), providing additional support for hypothesis 2. Thus, hypothesis 2 was supported by both analyses. The effect of growthT1 on growthT2 was positive but not statistically significant (b = .051).

Comparing Size Effects
The three first hypotheses were supported by our data concerning employment as well as sales. These findings suggest that there are mutual causal relationships between growth motivation and growth. Hypothesis 4 stated that growth motivation at T1 has a stronger impact on growth at T2 than has growth at T1. We tested this hypothesis by comparing the size of the regression coefficients of growth motivationT1 vs. growthT1 in model 6 concerning sales growth and in model 8 concerning employment growth. We found support for hypothesis 4 when employment was examined. In model 8, growth motivationT1 had a significantly larger effect than growthT1 on growthT2 (c2 = 9.11; p < .003). However, we did not find support for our hypothesis when sales was examined. For sales, growth motivationT1 had a larger coefficient than growthT1 on growthT2, but the difference was not statistically significant (c2 = .14).

Summary of Results
We find temporal stability of growth motivation, which is a prerequisite for growth motivation being a relevant predictor of growth. Our results also suggest an effect of growth motivation on growth. However, we also noted that past growth affected growth
3. As an additional test of the stability of growth motivation, we compared the mean values at T1 and T2. Employment growth motivation decreased by .17 SD and sales growth motivation decreased by .13 SD. Differences smaller than .25 SD are considered as very small and could be disregarded (Cohen, 1969). Thus, this test corroborates our previous findings.

May, 2008

449

motivation, suggesting mutual relationships between growth motivation and growth. All these results apply to sales as well as employment growth. In order to further test the relevance of growth motivation in studies of growth, we compared the size effect of growth motivation vs. past growth in the prediction of growth. The hypothesis that motivation has a stronger effect on firm growth than has past growth was supported for employment growth but not for sales growth. In sum, these findings suggest that growth motivation is a relevant predictor of growth and an important variable to include in studies of small firm growth.

Discussion
In this study, we have addressed the relationship between growth motivation and actual growth in small firms. The incentive for conducting this research is that recent research suggests that the relationship between the two concepts is more complex than previous empirical research suggest, calling for this type of research (Wiklund & Shepherd, 2003). Further, recent development of motivation theories indicates that the temporal stability of motives and the effect of previous behavior on motives affect the appropriate modeling of the relationship between motivation and behavior. The neglect to incorporate such constructs undermines our ability to more fully understand how motives affect our behavior and subsequent performance. More specifically, including notions of feedback loops and stability in motives over time is likely to lead to better models. By consecutively collecting data on both growth motivation and growth, correcting for sample selection and applying cross-lagged regression analysis, we have tried to test such a model of growth motivation and firm growth. Consequently, we were able to establish some important and interesting relationships between the constructs. We find that growth motivation has a unique impact on the growth of the firm, but that there are important feedback loops from growth to motivation. This provides support for the idea that the motivations of managers affect important firm outcomes such as growth. In other words, earlier research on the effect of motivation on firm growth appears valid, but it underestimated the importance of past growth in this process. Managers vary in their motivations to grow their firms, and those motivations affect growth achieved. Growth motivation, in turn, is partly affected by previous outcomes but remains relatively stable over time. This is an important result, as motivations have to be stable to be good predictors of behavior. Hence, growth motives are effective predictors of firm growth when they are stable over time.

Theoretical Implications
One important implication of these findings is that it makes sense to study motivation in the context of small firm growth. Small business managers do affect the growth of their firms by their intentional behavior. Another important implication is that while the bulk of previous research has relied on cross-sectional designs, the conclusions drawn—that motivation affects growth—were supported by our more careful analyses. It should be noted, however, that in the case of sales growth, past growth was an equally good predictor as was growth motivation. This suggests that at the least, past growth should be included as a control variable in models predicting growth in order not to overestimate the effect of growth motivation. This finding also leads us to speculate that there are some substantive differences between sales and employment growth. Sales growth reflects increases in output and is often used as a proxy for performance, while employment growth relates to increases in
450 ENTREPRENEURSHIP THEORY and PRACTICE

the stock of resources of the firm (cf. Penrose, 1959). The processes leading to sales and employment growth are quite different. Sales growth does not just happen because management wants it, but is determined by the market. Many different strategic choices can lead to sales growth, such as productivity increases, new product development, or new market entry. It is likely that such strategic considerations have long-lasting effects (if successful) and only gradually disappear in the face of competition. It also suggests that growth motivation only indirectly influences sales growth because relevant strategies have to be implemented to create sales growth. This explains our finding that sales growth in previous periods had a strong effect on future growth—just as strong as growth motivation. Conversely, employment growth is directly affected by the motivation of management. Small business managers have discretion to choose to add new employees or not and employees can be added to the firm more or less instantaneously. This would explain our strong effect of employment growth motivation and the nonsignificant effect of previous employment growth. Without doubt, the vast majority of studies of the relationship between growth motivation and growth has been concerned with how motivation affects growth. Our findings of feedback from growth on motivation suggest that there is ample opportunity to make a contribution to the literature by instead focusing on the reverse relationship, i.e., how outcomes affect future motivations and how these motivations change over time depending on whether intended outcomes materialize or not. For example, a natural extension of this study is to investigate heterogeneity of growth intentions among small business managers. How do previous outcomes affect the intentions of small business managers to grow their businesses in the future? The theory of planned behavior, modified in accordance with the findings of Bagozzi and Kimmel (1995), offers a theoretical framework for such an empirical investigation. It should be possible to incorporate previous work on the antecedents of growth motivation (Davidsson, 1989; Kolvereid, 1992; Wiklund et al., 2003) in assessing how attitude toward the behavior, perceived behavioral control, and social norms are influenced by past outcomes and how it affects intention. Our findings also add to the debate on the extent to which top managers influence the performance of their firms (see Bowman & Helfat [2001], and Carroll & Hannan [2000], for opposing views on the topic). Based on our findings, it appears that managers indeed do contribute to performance, at least in small firms. It appears that their aspirations to expand the firm affect subsequent growth. This suggests that motivation theories can be valuable in understanding the relationship between management and outcomes if feedback loops are accounted for. We also contribute to the literature on goal-directed behavior by testing the importance of motivation in a natural setting where the task is complex and the individual has limited control over outcomes. Most research in this domain examines simpler behaviors where individuals can more easily exert behaviors at will (e.g., driver compliance with speed limits [Elliott, Armitage, & Baughan, 2003] or choice of travel [Bamberg, Ajzen, & Schmidt, 2003]). Moreover, we use as dependent variable a relatively distal outcome (firm growth) rather than the actual behavior or a proximal outcome, such as the ability to solve a specific problem. This is an advantage when studying complex tasks where there are multiple ways of attaining the desired outcome (Campbell, 1988). The reliance of a narrow set of behaviors or proximal outcomes would probably have left out several possibilities of attaining the goal. Firm growth also has the advantage of being a measure of success or failure. The attainment of such outcomes has stronger effects on the individual’s subsequent motivation and learning than has the performance of specific behaviors (Silver et al., 1995; VandeWalle & Cummings, 1997). While we realize that the
May, 2008 451

study of complex behaviors under limited volitional control and distal outcomes introduces noise, our results suggest that it is possible to study such relationships in a meaningful way. We hope these findings will spur others to examine other relationships between the motivations of managers and firm-level outcomes.

Practical Implications
The practical implications of this paper’s findings are numerous. Small business managers with greater growth motivation are more likely to realize growth. This suggests that there is an opportunity for economic growth if small business managers’ growth intentions can be increased. Small businesses employ the majority of the workforce in most developed countries (e.g., Davidsson, Lindmark, & Olofsson, 1994; Storey, 1994), and small firms are of great importance to the development of these economies and the creation of new jobs (Carree & Thurik, 1998; Robbins, Pantuosco, Parker, & Fuller, 2000). Governments and others wishing to grow an economy need to understand that motivation plays an important role for the development and growth of small firms, and that measures to encourage the growth motivation of small business managers can have positive economic consequences. The importance of motivation has largely been overlooked in policy programs. So far there has been an overemphasis on implementing support programs that provide small businesses with resources that aim at increasing the ability for small businesses to grow, including training programs for small business managers and tax cuts. Implicit in most supportive programs is the assumption that if only small businesses had these resources and abilities they would grow. It may instead be possible for government to make growth a more attractive option for small business managers. For example, tax reliefs currently in place for very small businesses may make it less attractive for small business managers to expand their businesses beyond the point of receiving these relieves. Our results indicate that there are long-term effects of growth motivation because of feedback from previous outcomes. Growth motivation affects growth, which in turn has a positive effect on future growth motivation. This suggests that once a small business manager is motivated to try to expand the firm and is successful in doing so, his or her commitment to expansion will be reinforced. More generally, this finding suggests that the early outcomes of new firms operated by growth-motivated managers are of great importance. If they are able to achieve their growth targets, their motivation to further expand their businesses will be reinforced, leading to virtuous circles, with increased motivation and growth. Conversely, negative early outcomes are likely to lead to reduced growth motivation. Given that growth motivation is relatively stable over time but influenced by previous outcomes, the effects of such early outcomes are likely to be relatively long lasting. As noted earlier, however, motivation is not the only factor influencing the growth of small businesses. It is important that growth-oriented small businesses can access the resources they need at reasonable costs, and that growth opportunities are abundant in the economy. Furthermore, it is vital that they understand how to manage the firm through a growth process and understand the consequence of expanding a firm, most of which are positive. Furthermore, these results support the notion that intentional behavior has an impact on firm development, as suggested by psychological and strategy research, even if the impact is relatively small. Firm growth is not only the result of initial conditions, or
452 ENTREPRENEURSHIP THEORY and PRACTICE

follows an initial path. Both the behavior of the small business manager and the ongoing process of growth itself affect outcomes.

Limitations
Our data were composed of two different samples collected during different phases of the economic cycle. We find significant differences between the samples when modeling growth motivation. This suggests that to some extent, results might shift across samples, culture, and economic trends. However, our results hold in both samples which we interpret as an indication of robustness. We have also taken into consideration that sales growth and employment growth are not equivalent processes, and that they need to be studied separately. The similarities of results across these two measures further indicate robustness. However, there are important limitations to this study. While we believe that the results are likely to be generalizable to small businesses outside of Sweden, care must be taken in assessing country effects such as culture because growth motivation may, to some extent, have cultural roots. We relied on the CEO as the single informant. In cases where there are several owner-managers, this could lead to measurement error of growth motivation. Given that we used repeated measures, common method variance should not be a problem. Reliance on single informants should therefore not lead to spurious results because random measurement error attenuates rather than overestimates true relationships. In this article, we have only measured growth motivation and growth. However, owner-managers can have multiple goals that are interrelated. The specific mix of goals and how they relate to firm outcomes still needs more research. Although motives seem to be quite stable over time, some owner-managers will change motives depending on feedback and other information. For example, how will a growth-oriented owner-manager react to a major setback in the firm’s development, or how will an owner-manager motivated to remain at the same size react to a profitable opportunity that involves expansion in order to be exploited? Those are interesting areas for future research that are not covered in this article. The fact that we only measure growth motivation and growth also means that we cannot totally rule out effects of third variables and of intermediate variables such as strategy or behavior. We have tried to minimize the possibility of a third variable problem by adding a number of important control variables. Intermediate variables such as strategy and actual behavior would have been an important addition to our model because they are more proximal to the entrepreneurs’ or small business manager’s motives and abilities than are firm outcomes that a priori depend on a number of factors that the entrepreneur cannot control. However, firm outcome probably represents the most important form of feedback about the values of the strategies and behaviors initiated by the entrepreneurs, which leads us to conclude that our interpretation of results is valid.

Future Research
A better understanding of growth motivation and small firm growth could serve to more closely integrate current work in entrepreneurship and firm behavior. Future research should investigate in more details the interplay among motivation, strategy, firm operations, and firm performance. Studying these factors will give us a better understanding of how motives translate into behavior and how behavior affects performance and firm growth. This study has shown that small firm managers do affect the growth of their firm by their motivation, but how does this motivation translate into behavior, and which behaviors are more effective than others?
May, 2008 453

REFERENCES
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211. Ajzen, I. (1995). Attitudes and behavior. In A.S.R. Manstead & M. Hewstone (Eds.), The Blackwell encyclopedia of social psychology (pp. 52–57). Oxford: Blackwell Publishing Ltd. Ajzen, I. & Fishbein, M. (1977). Attitude–behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin, 84, 888–918. Ajzen, I. & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall. Anderson, G.A. (1991). How people think about causes: Examination of the typical phenomenal organization of attributions for success and failure. Social Cognition, 9(4), 295–329. Armitage, C.J. & Conner, M. (2001). Efficacy of the theory of planned behaviour: A meta-analytic review. British Journal of Social Psychology, 40, 471–499. Bagozzi, R.P. & Kimmel, S.K. (1995). A comparison of leading theories for prediction of goal-directed behaviours. British Journal of Social Psychology, 34, 437–461. Bagozzi, R.P. & Warshaw, P.R. (1992). An examination of the etiology of the attitude–behavior relation for goal-directed behaviors. Multivariate Behavioral Research, 27, 601–634. Bamberg, S., Ajzen, I., & Schmidt, P. (2003). Choice of travel mode in the theory of planned behavior: The roles of past behavior, habit, and reasoned action. Basic and Applied Social Psychology, 25, 175–187. Bandura, A. (1982). The psychology of chance encounters and life paths. American Psychologist, 37(7), 747–755. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall. Bandura, A. (1991). Social cognitive theory of self-regulation. Organizational Behavior and Human Decision Processes, 50, 248–287. Baum, R.J., Locke, E.A., & Kirkpatrick, S.A. (1998). A longitudinal study of the relation of vision and vision communication to venture growth in entrepreneurial firms. Journal of Applied Psychology, 83(1), 43–54. Baum, R.J., Locke, E.A., & Smith, K.G. (2001). A multidimensional model of venture growth. Academy of Management Journal, 44(2), 292–303. Bellu, R.R. & Sherman, H. (1995). Predicting firm success from task motivation and attributional style. A longitudinal study. Entrepreneurship and Regional Development, 7(4), 349–363. Bhidé, A.V. (1999). The origins and evolution of new businesses. New York: Oxford University Press. Bowman, E. & Helfat, C.E. (2001). Does corporate strategy matter? Strategic Management Journal, 22(1), 1–23. Brockner, J. (1992). The Escalation of Commitment to a Failing Course of Action: Toward Theoretical Progress. Academy of Management Review, 17(1), 39–61. Campbell, D.J. (1988). Task complexity: A review and analysis. Academy of Management Review, 13(1), 40–52. Carree, M.A. & Thurik, A.R. (1998). Small firms and economic growth in Europe. Atlantic Economic Journal, 26, 137–146.

454

ENTREPRENEURSHIP THEORY and PRACTICE

Carroll, G.R. & Hannan, M.T. (2000). The demography of corporations and industries. Princeton, NJ: Princeton University Press. Cohen, J. (1969). The statistical power of abnormal-social psychological research: A review. Journal of Abnormal and Social Psychology, 65, 95–121. Cohen, J. & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Conner, M., Sheeran, P., Norman, P., & Armitage, C.J. (2000). Temporal stability as a moderator of relationships in the Theory of Planned Behaviour. British Journal of Social Psychology, 39, 469–493. Davidsson, P. (1989). Continued entrepreneurship and small firm growth. Stockholm: Stockholm School of Economics, The Economic Research Institute. Davidsson, P., Delmar, F., & Wiklund, J. (2002). Entrepreneurship as growth; growth as entrepreneurship. In M.A. Hitt, R.D. Ireland, S.M. Camp, & D.L. Sexton (Eds.), Strategic entrepreneurship: Creating a new mindset (pp. 328–342). Oxford: Blackwell Publishers. Davidsson, P., Lindmark, L., & Olofsson, C. (1994). Dynamiken i svenskt näringsliv (business dynamics in Sweden). Lund, Sweden: Studentlitteratur. Delmar, F. (1996). Entrepreneurial behavior and business performance. Stockholm: Stockholm School of Economics, The Economic Research Institute. Delmar, F. (1997). Measuring growth: Methodological considerations and empirical results. In R. Donckels & A. Miettinen (Eds.), Entrepreneurship and SME research: On its way to the next millennium (pp. 199–216). Aldershot, U.K.: Ashgate. Delmar, F., Davidsson, P., & Gartner, W.B. (2003). Arriving at the high growth firm. Journal of Business Venturing, 18(2), 189–216. Eagly, A.H. & Chaiken, S. (1993). The psychology of attitudes (1st ed.). Fort Worth, TX: Harcourt Brace Jovanovich, Inc. Elliott, M.A., Armitage, C.J., & Baughan, C.J. (2003). Drivers’ compliance with speed limits: An application of the theory of planned behavior. Journal of Applied Psychology, 88, 964–972. Gartner, W.B. (1989). Some suggestions for research on entrepreneurial traits and characteristics. Entrepreneurship Theory and Practice, 14(Fall), 27–37. Heckman, J.J. (1979). Sample selection bias as a specification error. Econometrica, 47(1), 153–162. Kanfer, R. (1991). Motivation theory and industrial and organizational psychology. In M.D. Dunnette & L.M. Hough (Eds.), Handbook of industrial and organizational psychology (2nd ed., Vol. 1, pp. 75–170). Palo Alto, CA: Consulting Psychologists Press Inc. Kennedy, D., Lakonishok, J. & Shaw, W.H. (1992). Accomodating outliers and nonlinearity in decision models. Journal of Accounting, Auditing and Finance, 7(2), 161–193. Kolvereid, L. (1992). Growth aspirations among Norwegian entrepreneurs. Journal of Business Venturing, 7, 209–222. Kolvereid, L. & Bullvag, E. (1996). Growth intentions and actual growth: The impact of entrepreneurial choice. Journal of Enterprising Culture, 4(1), 1–17. Koys, D.J. (2001). The effects of employees satisfaction, organizational citizenship behavior, and turnover on organizational effectiveness: A unit-level, longitudinal study. Personnel Psychology, 54, 101–114.

May, 2008

455

Lind, M.L. & Zmud, R.W. (1991). The influence of a convergence in understanding between technology providers and users on information technology innovativeness. Organization Science, 2(2), 195–217. McCloy, R.A., Campbell, J.P., & Cudeck, R. (1994). A confirmatory test of a model of performance determinants. Journal of Applied Psychology, 79(4), 493–505. Miner, J.B., Smith, N.R., & Bracker, J.S. (1989). Role of entrepreneurial task motivation in the growth of technologically innovative firms. Journal of Applied Psychology, 74(4), 554–560. Miner, J.B., Smith, N.R., & Bracker, J.S. (1994). Role of entrepreneurial task motivation in the growth of technologically innovative firms: interpretations from follow-up data. Journal of Applied Psychology, 79(4), 627–630. Mok, A.L. & Van den Tillaart, H. (1990). Farmers and small businessmen: A comparative analysis of their careers and occupational orientation. In R. Donckels & A. Miettinen (Eds.), New findings and perspectives in entrepreneurship (pp. 203–230). Aldershot, U.K.: Avebury. Netemeyer, R.G., Burton, S., & Johnston, M. (1991). A comparison of two models for the prediction of volitional and goal-directed behavior: A confirmatory analysis approach. Social Psychology Quarterly, 54, 87–100. Nunnally, J.C. (1967). Psychometric theory. New York: McGraw-Hill. Nunnally, J.C. & Bernstein, I.H. (1994). Psychometric theory. New York: McGraw-Hill. Orpen, C. (1980). The relationship between expected job performance and causal attributions of past success or failure. Journal of Social Psychology, 112, 151–152. Ouellette, J.A. & Wood, W. (1998). Habit and intention in everyday life: The multiple processes by which past behavior predicts future behavior. Psychological Bulletin, 214(1), 54. Penrose, E. (1959). The theory of the growth of the firm. Oxford: Oxford University Press. Perugini, M. & Conner, R. (2000). Predicting and understanding behavioral volitions: The interplay between goals and behaviors. European Journal of Social Psychology, 30(5), 705–731. Robbins, D.K., Pantuosco, L.J., Parker, D.F., & Fuller, B.K. (2000). An empirical assessment of the contribution of small business employment to U.S. state economic performance. Small Business Economics, 15, 293–302. Rosenthal, R. (1991). Meta-analytic procedures for social research (Rev. ed., Vol. 6). Newbury Park, CA: Sage Publications, Inc. Shane, S., Locke, E.A., & Collins, C.J. (2003). Entrepreneurial motivation. Human Resource Management Review, 13(2), 257–279. Shaver, J.M. (1998). Accounting for endogeneity when assessing strategy performance: Does entry mode choice affect FDI survival? Management Science, 44, 571–585. Sheeran, P. & Abraham, C. (2003). Mediator of moderators: Temporal stability of intention and the intention– behavior relation. Personality and Social Psychology Bulletin, 29(2), 205–215. Sheeran, P., Orbell, S., & Trafimow, D. (1999). Does the temporal stability of behavioral intentions moderate intention–behavior and past behavior–future behavior relations? Personality and Social Psychology Bulletin, 25(6), 721–730. Silver, W.S., Mitchell, T.R., & Gist, M.E. (1995). Responses to successful and unsuccessful performance: The moderating effect of self-efficacy on the relationship between performance and attributions. Organizational Behavior and Human Decision Processes, 62(3), 286.

456

ENTREPRENEURSHIP THEORY and PRACTICE

Storey, D.J. (1994). Understanding the small business sector. London: Routledge. Thomas, K.M. & Mathieu, J.E. (1994). Role of causal attributions in dynamic self-regulation and goal processes. Journal of Applied Psychology, 79(6), 812–818. Triandis, H.C. (1977). Interpersonal behaviour. Monterey, CA: Brooks/Cole. VandeWalle, D. & Cummings, L.L. (1997). A test of the influence of goal orientation on the feedback-seeking process. Journal of Applied Psychology, 82(3), 390–400. Venkataraman, S. (1997). The distinctive domain of entrepreneurship research: An editor’s perspective. In J. Katz & R.H.S. Brockhaus (Eds.), Advances in entrepreneurship, firm emergence, and growth (Vol. 3, pp. 119–138). Greenwich, CT: JAI Press. Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Psychological Review, 92(4), 548–573. Weiner, B. (1992). Human motivation: Metaphors, theories, and research. Newbury Park, CA: Sage Publications, Inc. Weinzimmer, L.G., Nystrom, P.C., & Freeman, S.J. (1998). Measuring organizational growth: Issues, consequences and guidelines. Journal of Management, 24(2), 235–262. Whetten, D.A. (1989). What constitutes a theoretical contribution. Academy of Management Review, 14(4), 490–495. Wiklund, J. (1998). Small firm growth and performance. Jönköping, Sweden: Jönköping International Business School. Wiklund, J., Davidsson, P., & Delmar, F. (2003). What do they think and feel about growth? An expectancyvalue approach to small business managers’ attitudes toward growth. Entrepreneurship Theory and Practice, 27(Spring), 247–270. Wiklund, J. & Shepherd, D.A. (2003). Aspiring for and achieving growth: The moderating role of resources and opportunities. Journal of Management Studies, 40(8), 1919–1942. Wiklund, J. & Shepherd, D.A. (2005). Knowledge accumulation in growth studies: The consequences of methodological choices. Paper presented at the ERIM 29 Expert Workshop “Perspectives on the Longitudinal Analysis of New Firm Growth,” Rotterdam, NL. Yanovitzky, I. & Capella, J.N. (2001). Effect of call-in political talk radio shows on their audiences: Evidence from a multi-wave panel analysis. Journal of Public Opinion Research, 13(4), 377–397.

Frédéric Delmar is Professor of Entrepreneurship at EM Lyon, and Associate Professor at the Center for Entrepreneurship and Business Creation, Stockholm School of Economics. Johan Wiklund is Associate Professor of Entrepreneurship at Whitman School of Management, Syracuse University. Financial support was provided from Jan Wallander’s foundation, Knut and Alice Wallenberg’s foundation, Ruben Rausing’s Foundation and Sparbankernas Research Foundation. We thank Michael Frese, Gerard George, and Dean Shepherd for their comments on earlier drafts of this paper.

May, 2008

457