Lawrence F. Katz and Alan B. Krueger’s research on alternative work arrangements—especially their analyses that draw on the RAND-Princeton Contingent Work Survey and related comparisons to BLS Contingent Worker Supplement concepts—shaped much of the public debate about how common temporary, on-call, contract, and independent-contractor work had become in the United States. Staff Gauge treats that literature as a lesson in leveling definitions before reading a count: the headline share moves when the questionnaire, the sample frame, or the treatment of multiple jobs changes.
Definitions before magnitudes
In the line of work associated with Katz and Krueger, “alternative work arrangements” typically groups independent contractors, on-call workers, temporary help agency workers, and contract company workers—categories that echo BLS CWS wording while allowing survey experiments outside the official supplement calendar. The analytical claim is not that every short spell of paid activity belongs in that basket. The claim is that these named arrangements can be measured with parallel question modules and then compared across time or across survey vehicles.
That precision matters for secondary earnings. A person whose main job is a conventional wage post and who also sells occasional services may not answer as an independent contractor on a module that asks about the main job’s arrangement. Conversely, someone whose only attachment is contracting may be counted fully in the alternative share even if gross receipts are small. The arrangement label and the earnings volume are different staff marks on the same gauge.
Survey design and contested comparisons
Subsequent discussion of the Katz–Krueger estimates has emphasized how online or opt-in panels can differ from the CPS household frame in who responds and how carefully job spells are dated. When later BLS CWS releases showed smaller alternative-arrangement shares than some earlier non-official surveys, commentators disagreed about whether the economy had shifted, whether earlier panels overstated prevalence, or whether question order and reference periods differed enough to block a clean time series. Staff Gauge records that disagreement as contested: the literature does not settle on a single reconciled series without heavy assumptions.
Method papers in this area also stress multiple-jobholding. If a survey asks arrangement questions only about the main job, secondary activities remain off-stage. Modules that ask whether any work was performed through an intermediary platform add another layer—but only when those questions are fielded. Reading a 2010s paper as if it measured 2020s platform taxonomies imports vocabulary the original instrument may not have carried.
What the papers do not license
Katz and Krueger’s empirical work describes distributions of arrangements and, in related projects, wage and hour patterns conditional on those arrangements. It does not license claims that a particular reader can raise household income by switching arrangement types, nor does it supply a sequence of steps that produce earnings. Staff Gauge rejects copy that converts arrangement shares into personal forecasts.
The papers likewise do not replace tax or administrative evidence. Self-employment receipts on information returns, employer reports of temporary help payroll, or platform payment files measure different objects. Combining them without a crosswalk is a common source of inflated rhetoric about “the gig share of the economy.”
Editorial use
When Staff Gauge cites Katz and Krueger, the note names which survey vehicle and which definition of alternative arrangements is in play, and it flags when later official supplements disagree with earlier non-official estimates. The desk’s interest is the measurement seam between main-job arrangement labels and secondary earnings—not advice about employment choices.
Primary sources remain the authors’ published papers and the technical documentation of the surveys they analyze. Corrections: [email protected].