Claims about public goods— data edition

Data collection by government is good. It is necessary. It’s as reliable as the bureaucracy that Shepards its output.

I love how this article in Bloomberg, The BLS Can’t Be Replaced by the Private Sector, by Claudia Sahm slices and dices the public nature of data collection. But I wonder about her denial of its private creation. Let’s consider her claims.

To be clear, this is the way it should be: Private companies are not in the business of creating public goods, which is what economic statistics from the government are. They are free to users, transparent in their methods, protective of the privacy of individuals and businesses, and dependable. Most private companies that create statistics charge users for access. (Some share their results publicly as a form of marketing.) Many private companies share only limited information about their methodology to deter competition.

Private companies are in the business of making a profit, and to do that they need to attract customers. Some customers are more profitable than others. As a result, a company’s data will tend to reflect the needs of its clients — it won’t capture the economy as whole.

And yet— many privately created data sets are used openly in a transparent fashion.

Here’s a list summarized by Grok of data provided by private enterprises for public consumption:

Private organizations often provide data that can function as a public good—non-excludable and non-rivalrous, benefiting society broadly. Here are some examples:

1. Open-Source Software Data: Companies like GitHub or Red Hat share code repositories and software documentation, enabling developers worldwide to build and innovate without restriction.
2. Environmental Data: Private firms, like those operating weather stations (e.g., IBM’s Weather Company), provide real-time weather and climate data, which supports disaster preparedness and agricultural planning.
3. Health Data Aggregates: Pharmaceutical companies or research institutions sometimes release anonymized clinical trial data or disease prevalence statistics, aiding public health research and policy.
4. Geographic and Mapping Data: Organizations like OpenStreetMap or Google (via public APIs) offer mapping data that supports urban planning, navigation, and disaster response.
5. Educational Resources: Platforms like Khan Academy or Coursera provide free or low-cost educational content, democratizing access to knowledge.
6. Economic and Market Data: Financial institutions like Bloomberg or trading platforms sometimes release anonymized market trends or economic indicators, informing policy and research.
7. Scientific Research Data: Private research labs, such as those in biotech, occasionally share datasets (e.g., genomic data) that advance collective scientific knowledge.

These datasets, while often generated for profit, can be shared in ways that make them accessible and beneficial to the public, resembling public goods.

Furthermore, there are frequent examples of bureaucratic efforts being led astray by private subgroups. The temptation for biased numbers can occur in both sectors.

Can we think of any recent Federal agencies being dismantled?