본 연구에서는 조직후원인식과 정서적 몰입사이 관계에 대한 인과추정을 목적으로 한다. 조직후원인식과 정서적 몰입 사이 관계에 대한 기존 연구들은, 주로 사회교환이론 및 사회정서적 관점에 기반하여, 조직후원인식이 정서적 몰입에 영향을 미치는 것으로 여겨져 왔다. 하지만, 개인들은 본래 자신이 갖고 있는 정서적 몰입과 일치하는 수준으로 그들 환경에 대해 인식한다는 이론적 설명도 존재한다. 즉, 개인의 정서적 몰입이 그 개인의 조직후원인식에 영향을 미친다는 것이다. 이러한 상반된 이론적 설명이 존재함에도 불구하고, 기존의 연구들은 주로 횡단 연구방식을 취했기 때문에, 조직후원인식과 정서적 몰입사이의 인과추정을 체계적으로 수행할 수 없었다. 따라서, 본 연구에서는, 한 중견 의류업체로부터 얻은 이차 패널데이터에 대한 시차분석을 통하여, 조직후원인식과 정서적 몰입사이 관계에 대한 인과추정을 수행했다. 연구결과에 의하면, 조직후원인식과 정서적 몰입은 상호영향을 미치는 것으로 나타났다. 즉, 조직후원인식은 정서적 몰입에 영향을 미치며, 정서적 몰입 역시 조직후원인식에 영향을 미친다. 결과에 대한 몇가지 함의가 논의에서 제시된다.
The paper aims to make causal inferences in the relationship between perceived organizational support (POS) and affective organizational commitment (AC). Perceived organizational support is defined as employees’ global beliefs concerning “the extent to which the organization values their contributions and cares about their well-being” (Eisenberger, Huntington, Hutchison, & Sowa, 1986). Affective organizational commitment refers to employees’ “emotional attachment to, identification with, and involvement in the organization” (Meyer & Allen, 1997).
The existing research explained the relationships between POS and AC mainly on the basis of the social exchange theory and socio-emotional perspective (cf. Eisenberger et al., 1986; Lee & Peccei, 2007; Wayne, Shore, & Liden, 1997). That is, when employees perceive organizational support, they develop AC in order to reciprocate such support from the organization. Moreover, POS can play a role in satisfying individuals’ important socio-emotional roles, including needs for esteem, respect, and approval (Lee & Peccei, 2007). To the extent POS fulfills such socio-emotional needs of individuals, organizational membership and role status are integrated into their self-identity. Consequently, individuals develop AC. Thus, this approach assumes that POS affects AC. However, on the basis of logic derived from cognitive dissonance theory (Festinger, 1957), some authors (e.g., O’Reilly & Caldwell, 1981; Salancik, 1977) maintained that individuals’ prior attitudes lead to their perception about their organizational environments. According to this approach, individuals’ prior attitudes lead to their level of POS. That is, prior attitudes color individuals’ overall assessment of whether the organization is supportive. Although these competing theoretical explanations exist in explaining the relationship between POS and attitudes/behavior, because existing studies in this area mainly relied on cross-sectional research design, we still do not know which approach is more appropriate in explaining the relationship between POS and AC. In order to fill this research gap, the present study attempts to examine the relationship between POS and AC using two-wave panel data.
The relationship between POS and AC was examined using a two-wave panel data drawn from a small garment manufacturing firm that had approximately 200 employees. The first survey (T1) was conducted in 2005 May and the second survey (T2) was conducted six months later. On each occasion of measurement, the same questionnaire was sent. In total, the author collected a sample of 141 employees who participated in both surveys. Among them, 137 were usable.
The measures for all the variables used in the study were adopted from established measures whose psychometric properties were well proved. Collected two-wave panel data were analyzed using cross-lagged analysis. Some authors recommended cross-lagged analysis as a more correct way for examining causal inference between variables. That is, in order to examine the causal inference between any two variables (X and Y), we must examine the impact of the independent variable at T1 on the dependent variable at T2, after controlling the impact of the dependent variable at T1 on the dependent variable at T2. For variables X and Y, X is first regarded as the independent variable and Y is regarded as the dependent variable. Thus, after controlling the impact of Y at T1 on Y at T2, we should examine the impact of X at T1 on Y at T2. Then, Y should be regarded as the independent variable and X is regarded as the dependent variable. After controlling the impact of X at T1 on X at T2, we should examine the impact of Y at T1 on X at T2. Using these procedures, we can infer whether X causes Y or Y causes X. If X causes Y and Y also causes X, there is non-recursive relationship between X and Y. In examining two-wave panel data by cross-lagged analysis, the author used structural equation models with latent variables. Before examining the relationships between POS and AC, the author conducted preliminary analyses in order to check the reliabilities and factor structures of all the variables. The preliminary analyses showed that all the variables had acceptable reliabilities and factor structures.
The results of the main analyses showed that there exists non-recursive relationship between POS and AC. That is, as the social exchange and socio-emotional approaches suggested, POS affects affective organizational commitment. At the same time, as denoted by the principles of the cognitive dissonance theory, affective organizational commitment affects POS.
There are four implications for the results. First, there exists non-recursive relationship between POS and AC. That is, POS affects AC and, at the same time, AC affects POS. This suggests that there may be non-recursive relationships between POS and other organizationally desirable employee attitudes and behavior. POS has been shown to be related to a number of organizationally desirable employee attitudes and behavior, including turnover intention, absenteeism, in-role behavior, and organizational citizenship behavior. The existing research has proposed that POS affects these attitudes/behavior on the basis of the social exchange theory and socio-emotional perspective. However, the non-recursive relationship between POS and AC shown in the present study suggests the possibility that the relationships between POS and other attitude/behavior may be non-recursive.
Second, the non-recursive relationship between POS and AC extends the understanding of the antecedents of POS. The existing research suggests the positive work experiences such as fairness, supervisory support, and organizational rewards and job conditions as major antecedents of POS (Rhoades & Eisenberger, 2002). However, the results of the present study suggest that AC can be an important antecedent of POS.
Third, the recursive relationship between POS and AC offers increased understanding of the development dynamics for POS and AC over time, thereby providing insights into how to effectively manage both constructs. Both POS and AC have upward spiral or downward spiral relationship by affecting each other continuously. That is, low level of POS leads to low level of AC, and this low level of AC again leads to low level of POS. On the other hand, high level of POS leads to high level of AC, and this high level of AC again leads to high level of POS. Thus, there is downward spiral relationship between POS and AC, practitioners can focus on either POS or AC, or on both POS and AC in order to convert the downward spiral relationship into upward spiral relationship.
Fourth, the non-recursive relationship between POS and AC provides insights into how to effectively manage other important organizational phenomena associated with POS. For example, POS has been shown to be related to organizationally desirable employee attitudes and behavior such as in-role performance and attendance (Rhoades & Eisenberger, 2002). Thus, in order to make employees have more positive attitudes toward the organization or engage in more organizationally desirable behavior, practitioners can implement some managerial interventions toward POS or toward AC.
