Wage & Job Quality

Every statistic in this article’s macro data section — payrolls, unemployment, even JOLTS openings and hires — treats a job as a job. It doesn’t. A third axis runs underneath the whole paradox: what a job actually pays, for how many hours, and whether the price is even capable of clearing the market it’s supposed to clear.

What It Is

Two linked questions: is the “no qualified applicants” complaint actually, at least partly, a wage-offer problem in disguise — and is the composition of American job growth quietly getting worse even in periods when the topline numbers look fine?

How It Works

Start with what workers say they need. The NY Fed’s Survey of Consumer Expectations tracks the average reservation wage — the lowest wage a respondent would accept for a new job — and it hit a series high of $84,762 in March 2026, with the increase concentrated among men and college-degree holders: exactly the population most likely to be benchmarking against large-employer compensation rather than small-business compensation. In the same survey wave, satisfaction with wage compensation and promotion opportunities both hit their lowest levels since the series began in 2014, while the share of people actively searching fell — people less satisfied but searching less, consistent with discouragement rather than resolution.

That $84,762 figure only makes sense next to what small employers can actually offer, and the gap turns out to be structural, not a matter of will. The firm-size wage premium is one of the most robust, century-old findings in labor economics: firms with 500+ employees pay 30–50% more than firms with fewer than 25, holding worker type roughly constant — a gap comparable in size to the gender wage gap. The benefits gap compounds it: nearly 9 in 10 workers at medium/large businesses have access to employer medical coverage, versus roughly half at micro and small businesses. A small-business offer that looks numerically close to a large-employer offer on base pay is often meaningfully behind once total compensation is counted.

A fixed compliance or insurance cost spreads over fewer employees at a small firm, so the per-head burden is mechanically higher regardless of the owner’s intent — stacked on top of a broader economy-wide dynamic in which labor’s share of gross domestic income fell to 51% in Q1 2026, the lowest since 1947, while corporate profits’ share rose to 12.1%, the highest since 1950. This is a genuinely separate, deep policy question — regulatory design, market concentration, capital markets policy — worth its own article. The implication for this one is narrower but load-bearing: small businesses aren’t simply choosing to offer less. They’re paying more per employee for the same regulatory and insurance baseline, in an economy where the reward to capital is compounding faster than the reward to labor generally.

The compensation ceiling, in small businesses’ own data

In June 2026, only a net 28% of small business owners reported actually raising compensation — down three points from May and the lowest reading of the year — while only a net 17% planned to raise it in the next three months. In the same period, “labor cost” rose to the highest reading in NFIB’s survey history as the top business problem, precisely as “labor quality” complaints were declining. Read together with the 51%/84% “no qualified applicants” figures, this suggests a meaningful share of what gets reported publicly as a skills shortage is at least partly a wage-offer problem wearing a skills-shortage costume — small employers are increasingly cost-constrained on what they can offer, at the exact moment they report fewer qualified applicants. This directly echoes Peter Cappelli’s decade-old critique (see Screening & Exclusion) that employers under-report the wage-driven share of their own hiring difficulty, now with current-cycle data behind it.

One genuine complication guards against over-claiming that fixing wages alone would unfreeze the market. The Atlanta Fed’s Wage Growth Tracker distinguishes job stayers from job switchers; historically, switching paid a real premium, which is what drove healthy quits rates. For the first time since September 2010, that premium briefly inverted in 2025 — but by February 2026 it had returned, with switchers seeing 4.4% wage growth against 3.9% for stayers on a 12-month basis. Despite the incentive coming back, the quits rate has not recovered — still around 1.9%, sub-2% for nearly a year. Workers who could get a raise by switching mostly still aren’t switching, most plausibly because they lack confidence there are real, attainable openings to switch into. Wages are necessary, but on this evidence, not sufficient.

Underneath all of it sits a measurement gap the topline numbers simply don’t capture. The U.S. Private Sector Job Quality Index, maintained by University at Buffalo researchers, is built exactly to close this blind spot: it sorts jobs into “quality” (above-median pay and above-median hours) versus “low-quality” (below median on both) buckets, rather than counting every job as one unit. The index is down 12.8% since 1990 — in 1990 there were 94 high-paying jobs for every 100 low-paying ones; today that ratio is roughly 80 per 100. Sixty-three percent of all net jobs created since 1990 have been low-wage, low-hour jobs. Every payroll number discussed in Macro Labor Data — including June 2026’s +57,000 print, carried almost entirely by health care — deserves an implicit quality discount: a print dominated by aide-type job growth is not obviously equivalent, in job-quality terms, to one driven by professional or technical hiring, even though payroll counts treat them identically. U-6, the broadest official BLS underutilization measure (unemployed plus marginally attached plus involuntarily part-time), stood at 8.1% in May 2026 — roughly double the 4.2–4.3% headline rate reported the same period, capturing workers who show up as “employed” everywhere but are working part-time only because they can’t find full-time work.

Why It Matters