Is all turnover intent the same? Exploring future job preference and environmental considerations

ABSTRACT

This study approaches turnover intent in a novel way by incorporating both environmental and internal organizational factors together to create a more nuanced view of what drives turnover. The analytical focus is on senior-level employees in four agencies within the US Department of Health and Human Services. The findings show that internal organizational factors partially explain decisions to change jobs, but agency and time differences remain even after controlling for those factors. It also finds that the decision to leave government is driven by different factors than the decision to move to other jobs within government.

KEYWORDS:

• Turnover
• organization
• environment
• intent
• future job

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. Since this takes a subset of the sample, it is important to note that the findings are not based on a completely random sample of these types of employees in each agency, and response rates are unknowable.

2. The FedView survey asked respondents about race and ethnicity differently from year to year, yielding different categories from one survey to the next where some groups were not included in all years. This prevented me from using specific racial or ethnic distinctions for the full sample. I considered incorporating a minority/non-minority variable into my model, but decided against this approach after running tests on the FedView survey in years where racial and ethnic differences were broken into numerous categories. I found that turnover intent varied among many racial and ethnic groups, with non-minorities falling in the middle. Thus, a simple minority/non-minority categorization masked differences among subgroups. Because of this, I chose to exclude minority status from this model.

3. Recognizing that logit models may not be independent, the author also ran the model using a multinomial logit approach where the three categories were to stay in one’s job, desire a new federal job, or want to leave government. The findings were the same for this approach and the two binary models that were ultimately used for simplicity of interpretation.

4. Because the study uses pooled cross-sectional data rather than panel data, there is no way to identify if people appear in the sample multiple times. In future research, agency-level data from multiple years on actual turnover and agency characteristics paired with economic measures will allow for an assessment of the impacts of the political environment and economy on actual turnover. The agency-level approach creates panel data where more advanced time-series techniques can separate underlying time trends from other moving parts in the political and economic environment. However, this approach cannot capture the impact of many traditional variables that focus on individual perceptions and differences.

5. Additionally, the author conducted a series of sensitivity analyses to confirm findings by running the models independently for each agency to identify any anomalies in the data. There were slight variations, but nothing that challenged the validity of the approach and findings.

Additional information

Notes on contributors

Susannah Bruns Ali

Susannah Bruns Ali is an Assistant Professor at Florida International University. She previously worked in the US Department of Health and Human Services. Dr Ali’s research focuses on factors that influence public-sector employee career choices, with particular attention to the influence of the political environment on careerists’ choices.