Abraham and colleagues: drafting nontraditional work across survey and tax lines

When statisticians and labor economists try to measure work that does not fit a single long-term employer, they draft taxonomies: contingent, alternative, independent, platform-mediated, and other labels. Katharine Abraham and coauthors — including John Haltiwanger, Kristin Sandusky and James Spletzer in related programs of work — have written extensively on why household surveys and administrative tax data can disagree once those labels leave the drafting table. Draftwage treats their contribution as a lesson in category design for secondary and nontraditional earnings, not as advice.

Legal form versus economic activity

A recurring theme in this literature is that legal form (employee versus independent contractor, for example) is not identical to the economic activity of selling labor or services. Tax filings may record payments under contracting categories that household surveys classify differently when interviewers ask about “jobs.” Abraham and colleagues have emphasized that confusing the two produces undercounts or overcounts depending on which source is treated as truth. For secondary earnings, the problem sharpens: a small, intermittent payment stream may appear in information returns yet never be drafted as a second job on a household questionnaire — or the reverse, if respondents report informal activity that never hits a tax information form.

Why survey–administrative gaps appear

Comparative work associated with this research program documents gaps between survey-based estimates of nontraditional work and administrative measures built from tax records. Reasons discussed in the literature include reference-period mismatch, respondent classification error, incomplete coverage of platforms in older instruments, and the fact that some administrative series capture payments rather than employment relationships. None of those reasons authorizes inventing a single “true” side-income rate for a population without stating the instrument.

Method critique belongs beside any cited gap: administrative data miss some informal cash activity; surveys miss some short spells and misclassify others. Abraham-linked papers generally treat the gap as a measurement object, not as proof that one source should be discarded. Draftwage follows that discipline: when a sheet mentions a survey–tax disagreement, both limits stay visible.

Drafting implications for secondary earnings

For editors and analysts who care about secondary earnings, the practical drafting lesson is to name the unit (person, job, payment), the period, and the legal-versus-economic distinction before quoting a figure. Taxonomies that collapse platform tasks, classical second jobs and contracting into one “gig” label erase the distinctions Abraham and colleagues spent years clarifying. Conversely, taxonomies that are too narrow — only second payroll jobs — omit payment streams that administrative data show exist.

Contested territory remains: how often household surveys should be redesigned versus supplemented; how to treat online platforms in standard occupational and industry codes; and how much disagreement is tolerable before series are revised. Those debates are ongoing in the professional literature and are not settled by marketing language.

What this sheet does not claim

This sheet does not assert a preferred estimate of nontraditional or secondary work, does not tell readers how to classify their own activity, and does not promise any financial result. It summarizes a measurement agenda associated with Abraham and colleagues: draft categories carefully, compare survey and administrative lines, and keep legal form distinct from economic activity. Readers seeking personal guidance need regulated professionals, not an editorial board.

← Back to sheets