b'CHAPTER 5 | TRACKING THE IMPACT OF THE GREAT RECESSIONthe expected probability of working past age 65,FIGURE 5-2aAverage expected probability of working full time past age 62 for which increased from 39.6% to 49.5%.currently employed persons aged 55 to 60: 1998 to 2014HRS also assesses work plans by asking, Do55%you plan to stop working altogether or reduce50%work hours at a particular date or age, have you 45%not given it much thought, or what? The econom-ic environment influences retirement plans, but40%personal factors such as the amount of personal35%debt and characteristics of the individuals work30%environment are also strong predictors of work 25%plans (Szinovacz et al. 2013). 20%Flexibility in work hours15% 1998 2000 2002 2004 2006 2008 2010 2012 2014is associated with plans to FIGURE 5-2bAverage expected probability of working full time past age 65 for continue working currently employed persons aged 55 to 60: 1998 to 201455%Szinovacz et al. (2015) construct a model that captures several levels of influence, such as50%changes in the Dow Jones Industrial Average and45%the unemployment rate between 2006 and 2008.40%Mid-level factors include the size and degree of unionization, pension and health insurance ben- 35%efits, and opportunities to work part-time at the30%firms HRS participants work for. The model also25%includes individual characteristics like personal income and wealth, health status, tenure on the20%job, hours worked, and stressfulness and physical15%1998 2000 2002 2004 2006 2008 2010 2012 2014demands of the job. They find that unionization, having a DB pension, job stress, and onset of aSource: HRS 1998-2014. Men Womennew health condition are all associated with plans to stop working at younger ages. On the otherWhile work expectations generally predicttrend since 1992 in expectations of working past hand, flexibility in work hours is associated withactual work outcomes, it may be that the increasethe early (age 62) and normal (age 65) retirement plans to continue working. These effects may havein expectations of working past age 62 and pastages for Social Security. The 2008 wave of HRS changed as the recession evolved, which can beage 65 found over the recession will prove to besaw a big increase, followed by a decline back to investigated using data from subsequent wavesan overstatement of work outcomes. Figures 5-2athe trend line in 2010.in the HRS. and b show that there has been a slow upward 87'