Finding Accurate Causal Affects Of A National Supported Work

Better Essays
Paul Doxsee Econ 104 Finding Accurate Causal Affects Abstract An analysis was performed to see if the effect of a National Supported Work (NSW) Demonstration study was effective in increasing the earnings of a group of individuals. The analysis showed that, on average, those treated earned roughly $800 more than their counterparts three years later. Because of the randomization method used when selecting the participants, we can be confident that this estimate is causal. This randomization removes all biases from omitted variables as well as any selection bias, because the groups were systematically alike as shown in the analysis. Another analysis was performed using data from both the NSW study as well as the Panel Study of Income Dynamics (PSID). This analysis showed that using non-experimental data creates biased results because the two groups are not comparable. Furthermore, even the use of regressions in an attempt to correct this bias will not be successful and will still not result in causal results. Introduction The National Supported Work (NSW) Demonstration study was an experiment that took individuals with long-term employment problems and randomly assigned them to either receive or not receive a job training program. The goal of this program was to show that these job training programs work, in that they help people get better, higher paying jobs. The results of the study supported that. They showed that on average the earnings of the treated group after
Get Access