Global employee engagement fell to 20% in 2025, its lowest level since 2020, and Gallup's own modeling puts the cost at $10 trillion in lost productivity, roughly 9% of world GDP. Buried inside that headline is a regional number that matters more if your bench sits in South Asia: manager engagement there fell 8 points in a single year, the sharpest drop Gallup measured in any region. Most vetting checklists for a subcontracted development team ask about a portfolio, a pilot task, and an overlap window. None of them ask whether the manager holding your named point of contact together is one of the disengaged ones, and Gallup's own numbers say the odds of that just got worse.
What Gallup's 2026 report actually measured
Gallup's State of the Global Workplace 2026 report draws on 263,810 respondents in more than 140 countries, surveyed between January and December 2025, with 141,444 of them currently employed. It's the newest release in a tracking series running back to 2009 that has now collected more than 5.7 million responses. Global engagement came in at 20% for 2025, down from 23% in 2022 and the lowest reading since 2020, a second consecutive annual decline. Gallup estimates the gap between that number and a fully engaged workforce costs the world economy $10 trillion a year, about 9% of global GDP.
The steepest piece of the decline sits with managers, not the people they manage. Manager engagement dropped from 27% to 22% between 2024 and 2025, a 5-point fall Gallup calls the largest year-over-year manager decline it has recorded. Individual-contributor engagement barely moved over the same period, from 20% to 19%. Whatever changed in 2025 changed for the people running teams, not for everyone underneath them.
South Asia's managers took the sharpest hit
Regionally, South Asia posted the largest overall engagement decline Gallup measured anywhere: 5 points. Inside that number, manager engagement in the region fell 8 points, worse than the 5-point global manager drop and the steepest of any region in the report. Gallup frames the regional figure as driven primarily by India, tying it to organizational flattening and a slowdown in IT-sector hiring reaching mid-level and senior roles, some of it linked to AI adoption. The report doesn't break out Pakistan specifically, so treat that country-level detail as India's, not the whole region's, even though a Lahore or Karachi-based bench sits inside the same regional pool Gallup is describing.
Alongside the engagement number, the share of people in South Asia actually holding a manager title also declined, which Gallup reads as employers cutting management roles rather than just disengaging the ones who stayed. That matters because Gallup's own research on span of control found manager engagement erodes as a manager's team gets larger, an effect only manager talent and training can offset. Fewer managers over the same headcount is exactly the condition that model predicts keeps dragging the number down, not a one-year blip that corrects on its own.
Why a partner house's own engagement is your risk too
Our own guide to vetting a global development partner tells buyers to get one thing above almost everything else: a named, accountable person on the other side, not an account manager relaying messages. That's still the right advice, and Gallup's numbers are why it matters more this year than last. The person holding that role is, by definition, a manager, sitting inside the exact population whose engagement just fell twice as fast as anyone else's, in the exact region where the management layer itself is thinning.
A disengaged manager doesn't show up in a portfolio review. It shows up six weeks into the engagement: in slower responses to a blocked ticket, in a standup that quietly stops surfacing disagreement, in a written spec that stops getting written because nobody with the authority to insist on it is paying close enough attention anymore. None of that is a code-quality problem, which is why a pilot task built to test code quality won't catch it.
What this doesn't tell you
This is national and regional data, not a report card on any specific partner house. A team with a strong internal culture can sit well outside a regional trend, and a five-person bench you've already piloted and trusted for a year isn't a worse bet because of a country-level statistic that arrived after you signed. If you already have a relationship that's working, this isn't a reason to reopen vetting. It's a reason to ask two new questions the next time you do.
It also doesn't distinguish between two problems that look similar from the outside: a manager who has quietly checked out while staying on the roster, and one who has actually left, with a replacement not yet fully ramped. Gallup's survey measures the first. The second is a staffing question your contract's replacement clause is supposed to cover, and the two need different fixes. Treat a soft signal in a national survey as a reason to ask, not as a verdict on a team you already know works.
A worked example
Picture a 22-person Bay Area agency signing its second subcontract pilot with a Lahore-based partner house, a year after a strong first engagement with a different team there. The pilot task goes fine. The code review comes back clean, the demo lands on time, and the reference the agency calls says good things about the engineers on the project. What the agency didn't ask, because none of its vetting questions covered it, was how long the partner house's own delivery manager had held that role, how many other accounts he was carrying at the same time, and whether the house had lost any of its management layer in the past year.
Six months in, the answer showed up anyway. The manager who had been the named point of contact was by then managing three accounts instead of one. Standups compressed into status reads with no room for the disagreement that used to get raised there, and the written specs that made the first engagement work stopped arriving. Nothing about the pilot would have surfaced any of that, because a pilot tests the engineers, not the person managing them. Two questions in the original vetting call would have: how many accounts does our contact currently own, and has that number changed in the past year?
What this means for your agency
- Add manager span of control to the vetting call. Ask how many accounts or teams your named contact is currently responsible for, and whether that number is stable or growing.
- Ask the question again at renewal, not just at signing. A partner house's management layer can thin out well after your pilot proved the code was good.
- Don't treat a national statistic as a verdict on a relationship that's already working. Ask the two new questions rather than re-vetting a team that's earned trust.
- Watch for the early symptoms Gallup's data points to: standups that stop surfacing disagreement, written specs that stop arriving, a contact who's slower to respond than they were three months ago. Those show up before a formal complaint does.
We staff our own accounts the way we'd want a subcontract partner to staff ours: one named person per account, with the overlap hours and the writing habit that protect a distributed team from drifting into evening calls. If your current bench's management layer feels thinner than it did a year ago, that's worth a short conversation before it costs you a sprint, not after: tell us what's changed. And whether the fix is a better subcontract relationship or a dedicated team of your own, the disclosure and accountability questions in the economics of white-label development apply either way.