Bridging Generations: Understanding Work Values in Southeast Asia — Epoche B2
My Boss, My Intern, and Me: A Few Thoughts on Generations at Work Our new intern posted a one-line message in the team chat last Tuesday: 'Need feedback on the deck by 3pm today.' No greeting, no name, no question mark. I read it, thought 'fine, clear', and got on with it. My boss, who is fifty-eight, took me aside that afternoon and asked whether the intern was always that rude. Neither of them said anything to the other. Both of them said it to me. What follows is about that message: why one line of text produced two incompatible readings in the same Singapore office, and why the word everyone reaches for first — 'generation' — turns out to be the weakest part of the explanation. There are three of these age groups in our office, and I have stopped thinking of the differences between them as personality. But I want to be honest about the status of that claim before I make it, because the people who study this for a living are not convinced that 'generation' is a real variable at all. Why 'generation' explains less than it seems to This section deals with the objection, because everything after it depends on how much weight the word can carry. Costanza and Finkelstein, reviewing the workplace evidence in 2015, asked whether generational differences in things like job satisfaction, commitment to the employer and intention to quit are large enough to build management practice on. Their answer was that the measured differences are small — small relative to how much people within any one age group differ from each other — and that even those small differences cannot safely be attributed to the cohort a person was born into. The second half of that is the interesting half, and it is a structural problem rather than a matter of collecting more data. Suppose you survey a person and record three things: their age $a$ in years, the calendar year $p$ in which you asked them, and their year of birth $c$. Those three numbers are not independent, because $$c = p - a$$ always holds, exactly, for everyone. Sociologists call this the age–period–cohort problem, and Glenn's book on cohort analysis is the standard treatment of it. Its consequence is severe: because any one of the three quantities is fixed once you know the other two, no amount of statistical work can separate their effects without an assumption brought in from outside the data. So when our intern writes a message the way she does, three explanations fit the numbers equally well. It may be a cohort effect — she was formed by growing up with messaging apps. It may be an age effect — twenty-three-year-olds in any decade are brisk with senior people because they have not yet been burned. Or it may be a period effect — in 2020s office life *everyone* writes shorter messages, including my boss, who now sends me single words too. Nothing in the data chooses between these, and I cannot choose between them either. So what follows describes the three groups as they appear from a desk between two of them: observations about people who happen to share a decade of birth, not evidence that the decade caused anything. The three groups, and the events standing behind them Here is the composition of the office and, where I can identify one, the concrete experience that makes each group's expectations reasonable rather than merely characteristic. My boss's generation. Senior management, mostly in their late fifties and sixties. They started work when staying at one company for thirty years was the sensible plan, and several of them were here in 1997, when it briefly stopped being sensible and then became sensible again. That year is worth spelling out, because it is the one piece of their formation I can point to as an event rather than a mood. In the Asian financial crisis of 1997–98, capital that had flowed into the region reversed within months; Radelet and Sachs, writing in 1998, documented how quickly currencies and output fell across Thailand, Indonesia, Korea and Malaysia, with Indonesian output contracting by something in the order of thirteen per cent in 1998 alone. Singapore came through comparatively lightly, but 'lightly' meant a stalled economy and a resident unemployment rate that roughly doubled off a base under two per cent. What my boss's cohort watched, at the age when careers are made, was that the people who had banked long relationships with an employer were rehired first. Loyalty, hierarchy and turning up matter to them in a way that is not sentimental — they watched those things pay. Generation X, which is mine. We learned the office in both directions: the fax machine and the shared drive, the formal memo and the group chat. I think that is why we care about the result and not the channel — not from tolerance, but because learning a second set of conventions in mid-career made us experience both sets as conventions rather than as manners. It also means we spend a great deal of time translating, which is what this post is. The younger group. Most of our new hires. They want to know what the work is for, they expect feedback within days rather than at the annual review, and they use messaging apps the way my boss uses the telephone. In our Ho Chi Minh City and Jakarta teams this shows up more sharply still. On that last point I have to correct something I used to say, because writing it out showed me it was wrong. I used to claim that the demographic figures explained the difference: the ASEAN Statistical Yearbook, set beside the Department of Statistics Singapore's own Population Trends 2023, shows median ages in Vietnam and Indonesia running roughly a decade below Singapore's, whose resident median age passed forty during the 2010s. But a median age does not tell you anything about anyone's values. What it tells you is the mix: in an office recruiting from a population whose middle is around thirty, a much larger share of the room is young, so the younger group's conventions become the default rather than the exception. That is a co