In lieu of an abstract, here is a brief excerpt of the content:

Survey of Income and Program Participation T he book’s analyses draw data from a national survey—the Survey of Income and Program Participation (SIPP)—which includes ten panels and covers 1984 through 2003. This section provides justifications for using the SIPP and develops adjustments in order to improve the data and make it suitable for the analyses in this book. SIPP Data The survey design of the SIPP is a continuous series of national panels, with sample size ranging from approximately 11,000 to 35,000 interviewed households. The duration of each panel ranges from thirty to forty-eight months. The SIPP sample is a multistage-stratified sample of the U.S. civilian noninstitutionalized population. The 1984 panel was introduced in October 1983. For 1985 to 1993, a panel was introduced each year in February, which covers thirty to forty months. A forty-eightmonth panel was introduced in April 1996 and a thirty-six-month panel was introduced in February 2001. The SIPP collects monthly data every four months (a wave) by interviewing the original sampled adults and other individuals with whom they reside. The SIPP includes core questions asked in each wave and topical modules of specific topics asked in only certain waves. This book uses data from core questions on labor force participation and income and data from three topical modules: migration history, assets and liabilities, and education history. The migration history module asks whether each person in the household was born in the United States, and if born abroad, the country of birth and the year of arrival in the United States. The education history model collects information about respondent’s highest level of school completed or degree received, fields of study, and dates and places of receipt of high school and postsecondary degrees or diplomas. The assets and liabilities module asks about the ownership and amount of assets Appendix 281 and debts. Components of assets include residential home, vehicles, savings and checking accounts, stocks, mutual funds, bonds, real estate, business assets, and retirement accounts. Components of debts include mortgage, secured debts (such as home equity loans, car loans, and debts on stocks and mutual funds) and unsecured debts (such as credit cards and medical bills). The assets and liabilities module is included in ten of the twelve total SIPP panels (the 1988 and 1989 panels did not contain such a module). Five of the ten panels collected data on assets and liabilities for more than one wave (the 1984, 1985, 1986, 1996, and 2001 panels). Together, these ten panels provide eighteen cross-sections of data on assets and debts for fourteen years (1984, 1985, 1987, 1988, 1991, 1993, 1995, 1996 through 1999, and 2001 through 2003) because some panel years overlap. Quality of Wealth Data The availability of wealth, migration, and educational history data, together with its large sample sizes of pooled multiple panels and waves, make the SIPP the prime candidate for a study of immigrants’ wealth because no other national survey includes all of these attributes. However, the quality of the SIPP wealth data must be considered and compared with the wealth data provided by the Survey of Consumer Finances (SCF). The SCF is considered the most reliable survey data regarding the wealth distribution of the U.S. population. However, it lacks the migration and education history information necessary for studying immigrant wealth. In addition, its sample size is too small (fewer than 4,000 households) to allow studying small groups. The primary SIPP goal is to provide information needed for government program applications, including welfare programs, social security, Medicare, and Medicaid. It therefore pays less attention to the very wealthy and as a result the very rich are underrepresented. In addition, for confidentiality reasons, the SIPP topcodes components of assets. For example, the topcodes in the 1996 panel include $25,000 for saving bonds, $5,000 for own checking account , $375,000 for total equity value of properties, and $72,500 for total amount of joint interest-earning accounts. This topcoding, together with the undersampling of the very wealthy, leads to two limitations in the SIPP wealth data: the distributions of net worth and assets are incomplete , without an appropriate upper tail, and some components of assets relevant to the wealthy are left out, resulting in a smaller amount of total assets, overestimated shares of the middle portions, and underestimated shares of the upper portions. The two limitations have raised concerns over the use of the SIPP net...

Share