Preliminary Empirical Results--论文代写范文精选

对收益的影响在随后几年出现。转向其他异质性分析,估计影响总收益,尽管这些估计没有显著的不同。在我们的示例中,家庭收入低的中等收入相对于纽约是作为一个整体。下面的paper代写范文进行探讨。

Abstract 
Table 1 demonstrates the validity of the randomized design by comparing the characteristics of SYEP lottery winners and losers. Specifically, we run a “reduced form” OLS regression of characteristics of SYEP applicants on a dummy for winning the lottery and provider-by-year fixed effects. We examine outcomes in the year prior to applying to SYEP (total earnings, NYC government earnings, non-NYC government earnings, whether the individual has any job, whether the individual has a non-NYC government job, whether the individual is enrolled in college, total income of the family, and whether the individual participates in SYEP), and a number of demographic variables (number of family members, as well as dummies for race, gender, age, and whether the individual is a U.S. citizen). Consistent with the validity of the randomization, none of these eighteen variables is significantly related to treatment status. Moreover, for the outcome variables observed in the year before SYEP participation, we find insignificant estimates (with similar point estimates and small standard errors) in every other year prior to SYEP enrollment (Year -2, Year -3, Year -4, Year -5, and Year -6). 27 We also find that the probability that SYEP applicants match to the IRS data is balanced.

Heterogeneity 
In Table 4, we return to our two-stage least squares specification and investigate heterogeneity in the effect of SYEP among subsamples distinguished by different individual characteristics. 34 Rows A and B show the effect of SYEP enrollment broken down by eligibility for WOTC support of summer youth employment (i.e. those aged 16-17 living in an Empowerment Zone). 35 The effect of SYEP on contemporaneous earnings (Column 1) is positive and substantial in both groups. 

The effect on earnings in subsequent years (Column 4) is insignificant for those eligible for the WOTC, but the WOTC-ineligible group shows more substantial and significant negative effects of SYEP on total earnings in Years 1-4. 36 Turning to other heterogeneity analysis, the point estimate of the effect on total earnings in Years 1-4 is a bit less negative (though more statistically significant) in the below-median-family-income sample (Row C) than in the above-median sample (Row D), though these estimates are not significantly different. Note that in our sample, even those with above-median family income had very low median income relative to NYC as a whole. The effects are again similar between males and females (Rows E and F), with modestly more positive effects on total earnings among females than among males, though these effects are again not significantly different. The effect on earnings in Years 1-4 is much more negative among whites than among blacks, Latinos, and other races (Rows G through J). 

Effects on type of job 
We investigate the effect of SYEP on earnings in different industries. To simplify this exercise, we classify industries into two clusters: those in which the 2-digit industry is more prevalent (i.e. represents a greater percentage of total jobs) among SYEP-provided jobs than among jobs held by the control group (Cluster 1), and industries in which the opposite is true (Cluster 2). As shown in Appendix Table 1, Cluster 1 contains arts and recreation, camp (out of city), community/social service, day care/day camp, educational services, and healthcare/medical, while Cluster 2 contains the remaining jobs (in financial services, legal services, cultural institution, real estate/property, retail, science and technology, and“other”). 40 

This is a useful way of classifying the industries because it will allow us to examine whether SYEP led to a persistent increase in earnings in industries that SYEP typically places participants.41 Columns 1-2 of Appendix Table 9 show that in Year 0, earnings in Cluster 1 rise (as would be expected through a mechanical effect since jobs provided by SYEP are more likely to be in Cluster 1 than jobs outside of SYEP) and earnings in Cluster 2 fall. In subsequent years, these effects persist: SYEP participation tends to continue to positively (negatively) affect Cluster 1 (Cluster 2) earnings and employment.42 Columns 3-5 of Appendix Table 9 show the effects on earnings and jobs in the forprofit, non-profit, and government sectors. The table shows that SYEP strongly raises earnings in the non-profit sector in Year 0 and subsequently continues to raise earnings modestly in the non-profit sector through Year 4 (with similar results for the probability of having a job). Earnings in the for-profit sector are lowered by SYEP by around $100 per year in Years 0, 1, 2, and 3. Meanwhile, SYEP increases earnings in the government sector in Year 0 but has a modest negative effect on government earnings in Years 3 and 4.

Interpreting the Earnings Results 
While the negative effects on subsequent earnings are small relative to likely lifetime earnings, it is worth considering the reasons behind the arguably surprising result that SYEP participation decreases earnings among a young group with little prior work experience, even during the Great Recession. Our randomized design is well suited to determine the program’s causal impacts, but less equipped to determine the mechanisms that mediate these impacts. Thus, our exploration of the mechanisms is necessarily tentative: we can say whether the predictions of our hypotheses are consistent with the data, but we cannot determine with certainty which hypotheses are correct. With these caveats in mind, we turn to examining our suggestive evidence on mechanisms. 

Not mutually exclusive with the above or other explanations, SYEP could crowd out jobs that could have led to greater future earnings.45 As we discuss in Appendix 2, we find that groups that experienced greater Year 0 crowdout also experienced greater decreases in subsequent total earnings, as we would expect if crowdout of other experiences in Year 0 leads to decreases in subsequent earnings. Further, the subgroup analysis found more negative impacts for groups that were more likely to otherwise be working in Year 0 (i.e. older individuals, and those with a job in Year -1). Relatedly, SYEP decreases the probability that an individual continues working for a past employer (Appendix Table 11), raising the possibility that SYEP harms a participant’s career development by interrupting career development with a past employer. Appendix 2 also discusses other potential explanations, including income effects, substitutability of leisure across years, and signaling; the evidence on these mechanisms is mixed, though we cannot rule them out. 46 Finally, as discussed in the following section, we rule out the possibility that the decrease in subsequent earnings is a result of an increase in college enrollment.(论文代写)

51Due网站原创范文除特殊说明外一切图文著作权归51Due所有；未经51Due官方授权谢绝任何用途转载或刊发于媒体。如发生侵犯著作权现象，51Due保留一切法律追诉权。
更多论文代写范文欢迎访问我们主页 www.51due.com 当然有论文代写需求可以和我们24小时在线客服 QQ:800020041 联系交流。-X(论文代写)