Unemployment is 2.24% — near the lowest on record. 33.6% of workers are in vulnerable employment: own-account and unpaid family work, without a wage or a contract. Almost everyone has work. That is a different thing from almost everyone having a job.
The gap between 2.24% unemployment and 33.6% vulnerable employment is the whole story. Work is not scarce here; paid, secure work is.
Two labour figures that are both true and appear to contradict each other. Unemployment counts people looking for work and not finding it. Vulnerable employment counts own-account and unpaid family workers — people working, without a wage, a contract or any of what usually comes with a job.
Near the lowest on record. On this measure alone the labour market looks excellent.
A third of everyone working. Neither unemployed nor securely employed, and invisible to the headline rate.
A country can approach full employment and stay poor if the work is unpaid or informal. Reporting the first number without the second is the most common way to make this look better than it is.
The national poverty rate and the two international lines, at every year they were measured. This part is unambiguously good.
In 2023, down from 23.5% in 2015.
The World Bank's lower-middle-income line, comparable across countries.
Extreme poverty. Far lower, and the two lines moving differently is why quoting only one is misleading.
The distribution has been close to static for four decades, through every administration and every growth episode in that period.
Against 41.0 in 1985 — a move of -1.7 in 38 years.
Of all income, shared among 20% of the population.
4.6 times as much, among half as many people. That ratio is what the Gini number means in practice.
Five ASEAN economies at 2018, the latest year all of them have both measures. Comparing each country at its own latest print would put different years side by side and call the difference a gap.
At 42.3 in the comparison year. The Philippines is the most unequal of this group.
A middle-income country by every classification, with a lower-middle-income distribution.
Poverty falling while inequality holds means growth reached the poor — but reached everyone else at least as much.
The version this replaces charted agricultural wages by region, the gender wage gap, farm household income composition, rural income distribution and regional employment — using a 2015 wage, a 2002-03 farm income and a 2016 employment figure side by side as if they described one moment.
PSA's Family Income and Expenditure Survey carries all of it. psa.gov.ph sits behind a managed challenge that scripts do not pass, so no regional figure appears here at all.
Also PSA, and the specific figure the old page led with was a decade old when it was published. Nothing replaced it, because nothing open carries it.
The 2002-03 figures the old page used are three surveys out of date. A number that old is not wrong so much as no longer about the present, and presenting it beside 2016 employment made it look current.
One fetcher, four CSVs, via the shared helper in data/_lib/worldbank.py.
Poverty and distribution come from household surveys every few years. The survey table is published separately from the annual panel precisely so nobody joins the dots: 14 points across 38 years.
The $3.65 line must always catch at least as many people as the $2.15 line, and the bottom fifth must always hold less than the top tenth. Both catch a transposed column pair, which otherwise produces entirely plausible-looking output.
Nulls are never forward-filled. The national poverty series has four points; drawing it as annual would invent the rest.
Unemployment and vulnerable employment appear on the same chart by design. Either alone gives a badly wrong impression of the same labour market.
The ASEAN table uses the latest year every country has (2018), which is older than the Philippine latest and is labelled as such.
Eight assertions in checks.sql, and every figure bound to a query in facts.sql.
Every figure on this page traces to one of these, through a CSV in
data/ph-poverty/. Each is checked against its source query on every
build.
Poverty rates, Gini, income shares, unemployment and vulnerable employment — Derived from PSA household surveys; PSA's own portal is not reachable by script
I wrote a companion post — “Almost Everyone Works. A Third Have No Wage.” — in simple, everyday words.