Engineers' voluntary turnover: application of survival analysis.

AuthorMulla, Zubin R.

Introduction

Since the economic liberalization of the 1990s, there has been a flood of foreign direct investment by multinational corporations setting up and expanding their Indian operations (Holtbrugge, Friedmann & Puck, 2010). Indian companies too have responded to the liberal regime by expanding their businesses and their product range. Hence, in the last twenty years, there has been the creation of numerous opportunities for Indian engineers and managers. This has led to very high attrition rates in most Indian organizations (Gulati & Krishna, 2011; Nancheria, 2009). High attrition (employee-initiated separation) rate has implications for costs (or hiring and training), loss of organizational learning, and customer satisfaction (Koys, 2001). While most companies strive to limit their attrition it is also recognized that some amount of attrition may be necessary for the organization to enable churn in the talent and to bring in newer ideas (Kulshreshtha & Krishna Kumar, 2005).

Employee Turnover

Turnover research is often limited to studying job attitudes and their impact on turnover intention (Chen et al, 2011; Krishnan & Singh, 2010; Lai & Kapstad, 2009; Tett & Mayer, 1993); however, there are some studies, which have also looked at the impact of job attitudes on turnover (Cohen, 1993). In this paper, we investigate the causes of voluntary turnover for engineers in a large public sector corporation. The category of Indian engineers is a class worthy of study because of their pervasiveness in Indian organizations both public and private. This is evidenced by the fact that there have been many studies dedicated to Indian engineers (e.g., Das, 1998; Parikh & Sukhatame, 2004; Sarveswara Rao, 1972; Shanthamani, 1977). Since this study was based on archival data from company records, we were limited by those variables, which were available from the company's database. In addition to demographic variables such as age, gender, and social category, we studied the effect of location match, employee performance, and college ranking.

Demographic Variables

A meta-analysis of turnover has shown that gender and race have no impact on turnover (Griffeth, Hom & Gaertner, 2000). However, we have reason to believe that women engineers' experience of the workplace is quite different from that of males. For example, Parikh & Sukhatme (2004) have identified several issues faced by women engineers in India right across their career.

Similarly, Das (1998) in a study of engineers working for two Indian public sector undertakings found that women were paid ten percent less than men of equivalent qualifications and work experience were. A study of on-the-job search in urban India found that younger workers and married workers are more likely to engage in on-the-job search (Banerjee & Bucci, 1995). Hence, keeping in mind the Indian social structure and the high prevalence of patriarchy and caste related issues; we have included age, gender, and marital status as predictors of employee turnover.

In a collectivistic culture like India, interpersonal obligations to family members and social ties with one's community are an important consideration in evaluating one's employment. Because of this, mobility of executives is limited in India and employees prefer employment opportunities at or near their hometown. Hence, we included a variable to measure the extent to which an individual's posting is close to his or her hometown as a predictor of tenure.

Employee Performance

It has been widely recognized that not all employee turnover is the same. Since different employees have different performance, turnover functionality is dependent on the kind of employees who are leaving the organization (Beadles et al, 2000; Sturman et al, 2003). If high performers are leaving the organization, it is a cause for concern; however, if low performers are leaving the organization, it may be a healthy trend. There is some support for the fact that low performers are more likely to leave the organization and this relationship is especially strong when the organization has performance contingent rewards (Williams & Livingstone, 1994). Since the organization we are studying is a public sector organization and they do not have any form of performance contingent rewards at the level of graduate engineer trainee, it is difficult to predict the relationship between performance and turnover. Hence, employee performance is an important variable as a determinant of turnover.

Since the wave of economic liberalization in the early 1990s, there has been a huge demand for engineering and management talent in India. Responding to this high demand, there have been a large number of private engineering and management colleges, which have sprung up during the last two decades. The newer colleges however have not been able to reach the quality of the established government colleges such as the Indian Institutes of Technology (IITs). The government too has responded to the high demand for engineering talent by launching a number of new IITs in priority zones across the country. However, these colleges too are considered to be in a different league from the established (i.e., older) IITs. This has led to segmentation amongst engineering colleges and consequently among engineers graduat ing from these colleges. Some companies recognize this distinction in either of the two ways. First, they may recruit exclusively from the premium institutes and second they may provide differential grades, postings, or salaries depending on the prestige of the institute.

Method

We collected data from Bharat Petroleum Corporation Limited (BPCL), which is one of the largest public sector companies in India, ranked 225 amongst the Fortune Global 500 rankings and having a sales turnover of about USD 45 billion. We studied a sample of 2,141 engineers, which consisted of thirteen cohorts of engineers recruited from the year 2000 to 2012. Eighty percent of the engineers were recruited in the company as officer trainees/ management trainees for marketing unit and the rest as per the graduate engineering trainee (GET) scheme for refineries. The break-up of the number of engineers recruited from each batch is given in Table 1.

For each of the engineers we extracted the...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT