Staying Power

Upper Middle's “Workplace Relationship Survey" examined how credentialed professionals navigate dissatisfaction at work – specifically, whether they leave bad situations, try to fix them, or stay put and endure. The survey drew on Albert O. Hirschman's Exit, Voice, and Loyalty framework, which proposes that those are the only three responses available when to people who feel misaligned in their relationships. The findings suggest that exit is desired more often than attempted (50.2% say they would quit if they could), voice is suppressed less by fear than by capitulation (35.0%), and that what presents as loyalty is, in most cases, a commitment to a specific professional narrative more than a specific company or role. 

Industry and geography proved meaningfully predictive of both the desire to exit and the willingness to speak up. Income, counterintuitively, proved not not be predictive of much at all.

EXIT

Whether or not someone quits is largely a (though not entirely) a product of their confidence in the long-term value of the skillset or connection – but in an unexpected way. Workers who consider themselves irreplaceable express the lowest exit desire of any group: only 26.5% would leave their roles if they were financially independent, compared to 56.1% of those who consider themselves somewhat replaceable. Curiously, 58.8% of these self-identified irreplaceable workers believe they’re underpaid and 38.2% do not regularly confront their employer about that lagging compensation. These people have the highest confidence they could find a comparable role (though 20.6% isn’t very high), but remain loathe to use their leverage.

The workers most convinced of their institutional value treat that value as a reason to stay put. Work becomes a story about commitment and stability.

The serial quitter data produces a related irony. Respondents who have voluntarily left three or more roles are the most confident in finding a new job (only 19.4% not confident, compared to 30.8% among those who've never quit), and the most exit-ready in their current position (58.3% would leave if financially independent). They know they can do it. And yet… 80.6% feel underpaid in their current role – the highest underpaid rate of any quit-history group – and their silent loyalty rate (22.2%) is statistically indistinguishable from people who have never voluntarily left anything (30.8%). In other words, serial quitting increases workers’ confidence in the long-term value of their skills, but does not appear to produce greater contentment. The always wind up in more or less the same standoff and seemingly quit when they get bored of the specifics. Work is a story about the battle of talent against stasis.

The age data is where exit desire and exit viability pull apart most visibly. The desire to leave if financially independent climbs in a clean arc from 44.0% at 25–29, to 54.3% at 40–44, to a peak of 74.2% at 50–54. At that same moment, confidence in finding a comparable role within six months craters: 54.8% of 50–54-year-olds are not confident, roughly double the rate for 35–39-year-olds (27.5%). After 55, both numbers normalize – exit desire drops to 33.3% and not-confident falls to 33.3% – which suggests either acceptance or a change in the relevant calculation. What changes is the exit trigger: the correlation between age and citing non-work factors (health, family, geography) as the likely exit trigger is r=+0.200, the strongest age correlation in the dataset, and by 60–64 it's the plurality reason (47.1%), well ahead of pay (11.8%). A clear professional narrative has more and more value right up until the moment it has none at all.

VOICE

The most common reason respondents don't speak up is not fear of being fired (9.7%), not concern about their reputation (11.9%), and not the time cost (7.9%). It's the preemptive conclusion that speaking up won't produce results (35.0%). This is a psychological position, not a structural one, and it correlates directly with exit desire: respondents silenced by futility are 62.9% likely to want out if financially independent, compared to 43.3% for everyone else (r=+0.187). They are not passive, per se. They are people who have made a specific, dispiriting assessment and now behave accordingly. Naturally, that assessment is more common amidst stronger headwinds. Among Communications and PR workers, 50.0% cite futility as the reason they don't speak up. Among teachers and professors, 60.0% cite futility and 0.0% make requests of their employers with any regularity². Among IT workers, 12.5% cite futility and 33.3% make regular requests, the highest active voice rate in the sample. 

Financial runway data complicates the obvious explanation for this, which is that financial precarity correlates to silence. That’s a bit true: 45% respondents with less than one month of financial runway never make employer requests. But 42% of employees with 12 or more months of runway – the most financially insulated group in the sample – never make requests. Oddly, the most vocal group is the one with 6–11 months of runway – only 18.4% never ask – which suggests voice becomes a more viable choice when there’s asymmetric risk and a win is more likely than a loss. The most interesting data point on this is the voice frequency–underpaid correlation, r=−0.152, which suggests that more frequent asking is associated with feeling less underpaid³.

Still, the income-voice relationship inverts at the top of the bracket. The correlation between income and asking behavior runs r=−0.134: higher earners generally ask more. But among respondents earning $500k or above, nobody makes regular employer requests (0.0%), and 30.0% haven't made any in six months. This is not the same as passivity – it's a different register of voice, deployed at structurally specific moments (contract renewals, performance reviews, the pointed conversation) rather than as ongoing friction.

Interestingly, geography produces some of the sharpest variation here. Work cultures vary. New York and California, the two largest subsamples, land on opposite ends of voice frequency: New Yorkers speak are (didn’t see this one coming) roughly 30% less likely to give voice to a concern or complaint.⁴

LOYALTY

The stated reasons people stay are, in order: compensation (48.4%), stability (42.6%), schedule and flexibility (32.5%), benefits (27.1%), and lack of better options (19.9%). The last one is the most honest and almost certainly an undercount as it requires respondents to admit that the relationship is involuntary, which conflicts with the story most professionals maintain about their careers. Still, some 85.5% of the respondents less confident in finding a comparable role reported that they would leave if they were financially independent.

These are people who know the arrangement isn't working, don't believe they can improve it, and don't believe they can leave it. That is not loyalty. It’s retention as prelude to burnout. It’s also hell for employers. Almost 23% of checked-out respondents (7.9% of the sample) make monthly or even more-frequent employer requests. Checked out employees become extractive.

There are, of course, reasons to stay beyond resignation (a funny word in this context). Prestige is a small but efficient handcuff. Only 9.0% of respondents cite it as a reason to stay in their roles – but among those who do, 68.0% feel underpaid and only 32.0% would exit if financially independent, the second-lowest exit desire of any stay-reason subgroup. Tenure works similarly: the correlation between length of tenure and exit desire is r=−0.115, meaning longer tenure predicts less desire to leave, regardless of how the job is going. Respondents in their first year are 61.2% exit-ready (would leave if financially independent); by 8+ years that's down to 45.2%. But this isn't loyalty either – it's narrative dependency. At some point, workers only make sense as characters within one context. We’re sensitive to that.

Family background data underscores the fact professional decisions are often more about narrative than money. Respondents from wealthy family backgrounds make few employer requests and report being the most likely to stay (40%) even if financially independent (they would know). On the other hand, only 4% of respondents from precarious backgrounds say they’d stay if they were financially independent.

David Graeber called America a civilization organized around work as an end in itself rather than a means to one. This data suggests that the degree to which work functions as identity – rather than as a means to the end of eating – determines what people do when they become disappointed, disenchanted, detached, or, you know, just kind of bored.

CONCLUSION

The 75.5% underpaid rate in this survey is not a measurement of individual misfortune. It is a measurement of a structural condition that has been compounding since roughly 1979, when pay and productivity began to diverge – and which a credential has done relatively little to buffer. Since 1979, productivity has grown 3.5 times as fast as the compensation of a typical worker. Economic Policy Institute. Personal grievance (coupled with either exit or voice behaviors) is a rational response to an economy that has been underdelivering on that specific promise for two generations. This is worth holding onto when applying Hirschman's framework. He argued that loyal members who stay and endure, rather than exit or agitate, do so for one of two reasons: either they "are confident that things will soon get better," or they have "a special attachment to an organization." 

What this survey suggests is that white-collar workers are far more likely to have a special attachment to professional narrative than to the institution supporting that narrative by offering a title and trajectory. That is a different kind of loyalty than Hirschman described – call it narrative handcuffs. What many of us share is the desire to believe that we’ve spent our time on a project that makes sense. If and when our jobs cease to make sense, our professional decisions ensure that our resume do. It’s a consolation prize, sure, but far from nothing. We value it highly enough to let ourselves get lowballed.)