Technological improvements, as well as the rising pressure to reduce overhead costs, have pushed governments to experiment with new ways of providing public services.  One recent shift in the US government has been the rise of teleworking, both as a mechanism to reduce costs, and as a strategy to increase employee engagement and performance.  Indeed, since the 2010 Telework Enhancement Act, the Office of Personnel Management (OPM) has praised teleworking as a “strategic organizational change program” (OPM, 2012).  From the perspective of performance management, not only could teleworking improve how the US government provides services, it could also dramatically change who selects into (and stays in) the civil service.

While teleworking is described as a significant tool in improving organizational effectiveness, rigorous empirical evidence of this claim is sorely lacking.  Most studies compare reported job satisfaction of teleworkers to non-teleworkers, thus conflating sorting effects with the impact of teleworking itself.  Experimental evidence suggests short-term benefits of teleworking on an individual’s productivity.  Yet, from a public management perspective, it is not only important to use truly comparable populations when making claims about individual effects on productivity and engagement, it is also important to estimate the long-term effects of teleworking on organizational performance as a whole.

This study considers the effects of teleworking on the US Patent and Trade Office (USPTO), a pioneer in this field with almost 80% of eligible patent examiners working from home.  Using propensity score matching and changes in eligibility rules as an instrument, the study considers the effects of teleworking on employee productivity and attrition from 2005 to 2013. The analysis also considers who is most likely to telework, thus shedding light on how selection into the public sector might change as teleworking expands, and how the program affects those who stay behind.  Preliminary results suggest that while teleworking has large effects on attrition and may affect the characteristics of new hires, an increase in reported hours of work does not necessarily translate into overall improvements in agency performance. If teleworking is to expand as rapidly as the last few years suggest, we must re-consider how this large institutional change interacts with employee incentive schemes, in order to ensure that the anticipated benefits on performance are indeed realized.