The $1 trillion turnover figure is not inflated. It is barely sourced.

Provenance: Partly sourced. The figure reconstructs from its stated inputs, but the input it is most sensitive to carries no attribution. This describes how the figure is sourced, not whether it is true, and it is not a claim that anyone acted in bad faith.

Gallup's 2019 article opens by stating that U.S. businesses lose a trillion dollars every year to voluntary turnover. It never derives the figure. It gives a worked example at company scale, then states the national number separately, without connecting the two and without a source.

Reconstructing it

The article's own example uses a 100-person organisation at an average salary of $50,000, a 2017 U.S. turnover rate of 26.3% attributed to the Bureau of Labor Statistics, and a replacement cost of one-half to two times annual salary. It reports roughly $660,000 to $2.6 million a year, which checks out arithmetically.

Scale the same inputs to the country and the headline appears.

That is the trillion, to within rounding. The derivation was recoverable. It simply was never shown.

Which inputs are sourced

The quits count is cited to the Bureau of Labor Statistics and is not in dispute. The $50,000 average salary carries no source, though it is plausible against the BLS mean annual wage for that year. The replacement multiple, which the whole figure turns on, carries no footnote, link or attribution anywhere in the article. The piece contains no external citation for any of its financial figures.

Two things in Gallup's favour

This is not a number that was pumped up, and it would be lazy to write it that way. They used the bottom of their own range: at two times salary the same arithmetic gives about $3.8 trillion, so they published the version that made their case least dramatically. And the hedge is theirs, since the article calls the range a conservative estimate in the same breath as stating it.

The criticism is narrower and harder to dismiss. In an article whose entire argument is a dollar figure, the number that figure is most sensitive to is asserted without a source.

Where this joins up

That replacement multiple is the same kind of object as the 33%-of-salary rule traced elsewhere in this series. Both are rules of thumb about what replacing a person costs, and both are repeated far more often than they are sourced. Change that one multiple and the trillion becomes anything between half a trillion and four trillion. Nothing else in the calculation moves.

This one lands on us

Our own cost-of-turnover pages derive their published range from exactly the same one-half to two times multiple. That is 1,653 pages, which is 84% of everything published on this site.

An earlier draft of this page said roughly a fifth. That was not a stale figure, it was a fraction written without being calculated, in a page about numbers repeated without being checked. It is corrected here because leaving it would have been the exact failure this series documents.

What we do differently: the multiple is attributed to two named sources with live links, the arithmetic is printed on every page that uses it, and each of those pages states that the Bureau of Labor Statistics publishes the wage and not the cost. What is still true is that it remains a rule of thumb and a large part of this site depends on it. It would be incoherent to publish this page and not say so.

What the figure will and will not carry

It is reasonable for establishing that voluntary turnover is an economically large problem, as an order-of-magnitude framing stated as one, and as a prompt to calculate your own cost from your own data.

It will not carry precision of any kind, since the range implied by Gallup's own multiple spans roughly one to four trillion dollars. It will not support comparison across years, because the salary and the multiple are fixed assumptions rather than measurements. It should never be attributed to the Bureau of Labor Statistics, which supplied the quits count and nothing else.