2 min read
dataviz-ph: did public-works money go where the poverty was?
- civic-tech
- data-viz
- case-study
PHP 5.04 trillion in DPWH contract awards went out between 2014 and 2024. Plot that spend against poverty across the provinces and the cloud has no shape. Plot poverty against economic output and a clean diagonal appears. dataviz-ph is the animated chart that lets you watch both, then decide for yourself.
What the data shows
Joining procurement with outcomes on one set of axes lets you ask whether public money tracked need. Across 81 provinces plus Metro Manila:
- DPWH spend per capita versus poverty incidence: rank correlation about +0.12 (p = 0.268). For all-government spend it is about -0.07 (p = 0.548). No pattern either way.
- Per-capita GDP versus poverty: about -0.53 (p < 0.001). Wealthier provinces are clearly less poor.
- Cumulative DPWH spend per capita against the change in poverty is the one weak signal: rho about -0.25 (p = 0.026), and still only association.
So the spend-versus-poverty cloud is shapeless while GDP-versus-poverty is a downward line.
How it is built
A static ECharts site, no backend. Every figure is computed by the pipeline and recomputed live in the page from public sources (PhilGEPS contracts, PSA poverty and GDP, DPWH tags), not hand-typed. Each preset finding carries a permutation p-value (10,000 seeded shuffles). Pick any non-preset pair of indicators and the page computes a live Spearman correlation with a 1,000-shuffle p-value at the displayed year.
What it is not
These are contract awards, money committed, not verified disbursement, and the gap between the two is exactly what province-level reporting does not publish. Correlation, not causation.
All figures derive from public records (PhilGEPS, PSA, DPWH). dataviz-ph computes statistical indicators only; specific claims require independent corroboration.