2c. Interpret the figure below (Figure 8). What does it show about earnings? 7500 Earnings ($) 7000 -.3 -2 -.1 GPA distance FIG. 8.-Quarterly earnings by distance from GPA cutoff. Lines are fitted val- ues based on the main specification. Dots, shown every .05 grade points, are roll- ing averages of values within .05 grade points on either side that have the same value of the threshold-crossing dummy.2d. Explain the findings from the Table 5 below. What do the results show for the first two columns (Main and Controls). Bonus: Explain what columns 3 and 4 are showing. Table 5 Earnings Effects 8-14 Years after High School Completion Main Controls BW=.5 BW=.15 Local Linear Reduced-form estimates: Above cutoff 372" 36649 409* 4794 410#4 (141) (130) (154) (198) (147) Instrumental variables estimates: FIU admission 1,593* 1,575* 1,665* 1,700 2,001* (604) (584) (645) (621) (696) Years of SUS attendance 815+ 7924 8335# 977*$ (276) (262) (271) (305) (306) BA degree 6,547 6,442 7,366# 10,769 5,958*4 (2,496) (2,411) (2,998) (5,726) (2,024) N 6,542 6,542 9,659 3,294 6,542 NOTE. FIU = Florida International University; SUS = State University System; BA = bachelor's degree. Standard errors are clustered within grade bins. The p-values are calculated using a clustered wild bootstrap- procedure described in Sec. III and app. B. The dependent variable in each regression is average quarterly carmings in 2005 dollars. The "BW=.15" specification uses observations within .15 grade points above and below the cutoff and allows for a linear trend in distance from the cutoff. The "BW=.5" specification uses observations within the .5 grade points on either side of the cutoff and allows for a quartic polynomial in distance from the cutoff. The "Local Linear" specification is identical to the main specification, but it allows for linear slope terms in distance from the cutoff that differ above and below the threshold. Significant at the 10% level. Significant at the 5% level. Significant at the 1% level