The mechanism in Bersin's account is administrative rather than emotional, which is why it has outlasted every management theme he has watched come and go in almost thirty years of studying HR [1]. Risk that used to sit with the employer moved onto the household: retirement through the 401k, health costs through savings accounts, and career progression through community college, apprenticeships and self-directed growth [11]. Behind that he reads a steady retreat of unions, collective bargaining and legislative support for workers [12], and argues the internet and AI made it ordinary to hire contractors anywhere while instrumenting the workplace to see who is delivering [13]. An engagement programme is not arguing with a mood; it is arguing with plumbing that was re-laid over three decades.
Take the PwC pair at face value and it will not close. Hold frontline trust at exactly 50% and the other 28% of the workforce would have to register about 111% for the whole to average 67% (0.72 x 0.50 = 0.36; 0.67 - 0.36 = 0.31; 0.31 / 0.28 = 1.11) [3]. Different instruments or different populations would explain that, and the piece does not say which. If your own engagement survey averages across a workforce this uneven, it will hand you the same comfortable middle and hide the same distribution.
A skeptic would say trust indices measure mood, and that mood recovers. On the leadership series the objection has some purchase, since the move Bersin reports is six points, a relative fall of roughly a quarter over six years [4], on figures he does not source line by line. The slower markers are harder to wave away. Company tenure on the S&P 500 is about two thirds shorter at the midpoints of the ranges he gives, 12.5 years down to 4.5 [2]; a career layoff average of 2.5 that has more than doubled implies something under 1.25 three decades ago [1]; and the federal minimum wage has held at $7.25 since July 2009, nearly 17 years as of 2026 [10]. Those are structural readings, and they have all moved one way.
The tradeoff a retention budget faces is which of two instruments it funds. Loyalty instruments assume the employee expects to still be there in five years, and Bersin reports more than 65% of workers already running a side hustle [5]. Portable instruments, meaning pay, schedule control and skills that travel, hold people while lowering the cost of their exit, and that is the genuine price of switching rather than an objection to be managed off the agenda.
Bersin's own reading is not gloomy. He treats AI as a positive for workers [14] and expects the effect to sort by occupation, with nurses, machinists, plumbers and other skilled trades holding up while accountants and administrative staff carry the worry [15], even as he notes the Michigan sentiment measure at a 60-year low [9]. That occupational split is the planning question for this quarter, because reassurance and re-pricing are different budget lines and only one of them keeps a machinist. Who ultimately absorbs the risk of a self-managed career is the question for the decade, and it will be settled well outside any single HR plan, which is the reason to make this quarter's promises ones the structure can actually keep.
