The Four Numbers Were Right. Nobody Had Opened The File.

This page had four numbers on it. I went to check them, expecting the worst.

They were all correct.

41,544 households. An average income of ₱247,556. A middle income of ₱164,080. An inequality score of 0.444.

I opened the actual survey file and worked them out again. 41,544. ₱247,556. ₱164,080. 0.4438.

So somebody had read them off the front of the dataset. Fair enough. But nobody had ever opened the file, and every chart on the page was still made up.

The File Opens Fine

The government agency that runs this survey blocks programs from its website. That is why so much of my other work stops at the national level.

But a copy of the public file sits on Kaggle, and it hands it over without asking who you are. The old page even linked to it. Nobody clicked.

41,544 households. 60 columns. Income, and where every peso went.

The Oldest Rule In The Book

Here is the thing I most wanted to check.

There is a very old idea in economics: the poorer you are, the bigger the share of your money that goes on food. It is called Engel's law, from the 1850s.

So I split the households into ten groups by income, poorest to richest, and asked each group what share of income went on food.

The poorest tenth: 59.62 out of every 100 pesos.

The richest tenth: 21.36.

And it falls at every single step in between. Ten groups, nine steps down, not one wobble.

A rule written down in Germany in the 1850s, holding perfectly in Philippine households in 2015. I find that genuinely lovely.

Why I Use The Middle, Not The Average

Look again at those two numbers from the top. Average ₱247,556. Middle ₱164,080.

The average is 1.51 times the middle. That is a big gap, and it happens for a simple reason.

The richest household in the file reported ₱11,815,988 for the year. One household like that pulls the average up for everyone. It cannot pull the middle.

So when someone quotes the "average Filipino household income", check which one they mean. The two describe quite different families.

Who Holds What

The richest tenth of these households held 34.32 out of every 100 pesos of income in the whole sample.

The poorest tenth held 2.16.

The top 1 out of 100 households — about 415 of them — held 8.14 on their own.

Two Gaps Worth Naming

Farming households earned ₱105,157 at the middle. Everyone else earned ₱182,696. So a farming family takes home a bit over half.

And by place: Metro Manila's middle household earned ₱312,275. In ARMM it was ₱111,394. Nearly three times.

The Catch, And It Is A Real One

These are 41,544 households that were surveyed. Not the country.

Proper surveys come with weights — numbers telling you how many real families each surveyed family stands for, so you can scale up. This public copy does not include them.

So everything here describes the sample. I have not scaled anything up to the whole Philippines, and if I did it would be wrong.

It is also from 2015. A decade ago. The pesos are 2015 pesos and I have not adjusted them.

Smaller claims. But every one of them came out of the file.