17.11.2014 Views

Survey Estimates of Wealth - Mathematica Policy Research

Survey Estimates of Wealth - Mathematica Policy Research

Survey Estimates of Wealth - Mathematica Policy Research

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

last day <strong>of</strong> November 1998 through the last day <strong>of</strong> February 1999. 2<br />

The newly expanded HRS<br />

was conducted in 1998. To assess the SIPP’s capture <strong>of</strong> change in wealth over the 1990s, we<br />

also present limited estimates comparing the SIPP and the SCF in late 1992 and early 1993, and<br />

we include additional SIPP estimates from early 1995 and from waves 3, 6, 7, and 12 <strong>of</strong> the 1996<br />

panel, which are centered on early 1997 and 1998, mid-1998, and early 2000.<br />

The remainder <strong>of</strong> this report is organized as follows. Chapter II compares recent estimates<br />

<strong>of</strong> wealth from the SIPP with estimates from the SCF and the PSID. Chapter III extends this<br />

comparison to a number <strong>of</strong> subpopulations and includes a comparison <strong>of</strong> SIPP and HRS<br />

estimates <strong>of</strong> wealth among the persons born before 1948. The findings presented in these two<br />

chapters provide the basis for a more detailed review <strong>of</strong> factors that account for the SIPP’s<br />

general underestimation <strong>of</strong> levels <strong>of</strong> wealth and its misrepresentation <strong>of</strong> at least some aspects <strong>of</strong><br />

recent trends. The results from these analyses are reported in Chapter IV. Chapter V presents<br />

the results <strong>of</strong> an application <strong>of</strong> econometric modeling to decompose the differences between<br />

SIPP and SCF estimates <strong>of</strong> wealth and to test an approach to adjusting the SIPP microdata to<br />

more closely approximate the distribution <strong>of</strong> wealth in the SCF. Finally, Chapter VI presents a<br />

number <strong>of</strong> recommendations regarding the ORES’s continued use <strong>of</strong> SIPP wealth data, including<br />

strategies to adjust the SIPP wealth data and, thereby, improve its resemblance to the wealth data<br />

from the other surveys, and both interim and longer term strategies that the data producer—the<br />

Census Bureau—could employ to improve the quality <strong>of</strong> the SIPP wealth data.<br />

2 The Census Bureau interviews one quarter <strong>of</strong> the SIPP sample—a “rotation group”—each month. The<br />

interviews for each wave <strong>of</strong> data collection, therefore, are distributed across four months.<br />

10

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!