As I read it, the data about emissions specifically is self-reported by facilities:

> We reached out to each of the top 200 facilities (ranked by the level of nearby cancer risk) to ask them if their emissions reporting was accurate — and if not, whether they would resubmit 2014-18 data to the EPA. Of the 109 companies that responded to us, 71% confirmed that their reported emissions were correct, and 29% noted errors, which we asked them to correct.

I live in the western hemisphere's largest integrated industrial complex (Freeport, TX integrated with the eastern edge of Houston as well). Note that Freeport, TX has ZERO state or federal EPA VOC analyzers which can actually detect which chemical is leaking. They can only detect "this amount of something with either sulfur or N-O bonds -- no clue what though!". Completely fucking useless for an area which manufactures something like 15-20% of all USA domestic chemicals.

The entire east side of Houston metropolitan area has only 3 air quality monitors which test for these kinds of chemicals. During huge major events like the ITC fire, they often show no increased pollution at all. I lived next to leaks every day and because I worked in the plants I knew the smells - one day acrylates, next day thiols, next day hydrocarbons, etc. But the 3 monitoring sites over 10 miles from me showed nothing at all.

Here is the one "correct" monitoring station near the chemical plants of Houston: https://tceq.maps.arcgis.com/apps/webappviewer/index.html?id... but several of its analyzers are often offline/broken/pending maintenance. Here's a map of all the other ones: https://tceq.maps.arcgis.com/apps/webappviewer/index.html?id... Generally single/dual color dots mark "not-useful" monitoring sites which might measure only PM2.5 or Ozone, for example. The 4+ color dots are generally useful.

The data used by ProPublica is far worse than the data here -- because what ProPublica used was "self-reported" data from the chemical plants. But living next to them and working in them, I know that many leaks are never reported and many leaks are never even known internally! Our government's data collection is a travesty. ProPublica couldn't use the real air quality measurements because having 2-3 points across 1000 mi^2 is completely useless for the wind models they wanted to apply to the problem.

We don't actually have any data. The government is failing us. They need to spend about $1 million per air monitoring station and build them along the perimeters of each plant so that leaks can be assigned to the offending companies, and they need to be built near housing so that we know how families are being affected.

ITC fire which blanketed houston's sky in smoke: https://abc13.com/deer-park-fire-2019-itc-houston-air-qualit...

The companies submitted bad data, and it was not until ProPublica visualized what the data were saying that it became obvious.

If anything it shows the government is not providing enough oversight on the data provided.

> 29% of facilities ProPublica contacted said the EPA’s data on them was wrong.

I don't think it's necessary to point out why this is not very convincing

> And why didn’t the government pay for an interface and analysis that the public could easily use?

Because "the government" -specifically, the executive branch- has its upper leadership torn out and replaced with a completely new bunch of people with hugely different ideology if not an explicit political interest in dismantling the work of their forebears every 4 or 8 years. This is most obvious with the recent Trump admin, but applies in degree to every single administration.

There used to be an internal, cross-disciplinary office that did API and visualization work for accountability reasons (unfortunately I forget the name). They did great work. It was dismantled and 90% of it was not replaced. The 10% that was replaced consists of (1) internal work (EIA being a good example, DOT being a bad one) that is driven entirely by the good will of employees who are fighting tooth and nail for any amount of budget to expose any amount of the work they do. And (2) external work by any number of massively wasteful contracting programs, like SBIR, where I used to work.

Reagan's SBIR program mandates a certain amount of money to be spent on tiny (100k-1m) contracts which are either useless or unused, because there's no way to actually convert anything to a long-term status. Most of that money goes to defense, but make no mistake that the work there is also completely useless. The whole program should be lit on fire.

> I think people believe their tax money could be spent more wisely and competently.

It can, if people stop voting for republicans who intentionally destroy the only functional programs and then divert the money to completely worthless contractors in the name of "free markets".

> Because "the government" -specifically, the executive branch- has its upper leadership torn out and replaced with a completely new bunch of people with hugely different ideology if not an explicit political interest in dismantling the work of their forebears every 4 or 8 years.

And what's worse is that one political party has been running a decades-long campaign to literally make government worse in order to get people on board with dismantling big sections of it. It's not even a conspiracy theory, it's an actual part of their public campaign platforms that people deliberately vote for.

It's hardly only the republicans blowing government funds on useless programs, Obama's team did the same with bogus 'clean coal' funding funnelled into public-private partnerships with coal outfits (to the tune of ~$8 billion via the DOE), gave loans to the renewable energy companies with the worst products on the market via insider connections (Solyndra), etc. It's an across-the-board problem.

If we really wanted to do large-scale infrastructure projects that were successful, like nationwide high-speed-rail, maybe we should study how China does it.

Conversely 70% of the data is good and valuable.