A proposal to fix the sad state of government statistics in the US. Many of the problems discussed are related to relying on surveys (with dropping response rates), high cost of collecting data and low trust in government.
I read this book in order to get fresh ideas how to improve collection and dissemination of government statistics. However I found few novel ideas. In my part of the world (i.e.Nordic countries) we solved most of the statistical problems the book seeks to address decades ago simply by improving government registers to near 100% coverage, nearly always up to date. This makes most of the costly surveys and censuses the US rely on superfluous. We get much more reliable data at a lower cost.
***I was granted ARC of this via Netgalley from the publisher.***
Today, we are generating more data than ever before. How the government collects data and puts it to use is important. In the book, Democratizing Our Data: A Manifesto by Julia Lane examines the current state of how the government collects data and how it can change in the future for the better. The author argues that the government now does not have the right tools to collect data in an accurate and timely way. Since decisions that impact policy are made with this data, the author argues that the government should be innovative in the way it collects data and make it more efficient while protecting the privacy of the people it has collected data on. I think that the author has presented a good argument and also presents a possible answer to the problem. There are times when the author goes over the same material again that can bog down the reading a little but otherwise a good read. I would recommend this to anyone interested in data science.
The only reason I gave this book 2 stars instead of 1 is because it exposed me to an issue I was unaware of. Otherwise, it is a self-promoting work that spends half the time focused on the author’s contributions to the issue, rather than the actual problem. This should have been a white paper, not a book. It’s a product of an egotistical author that robbed me of 4 hours of life I’ll never get back.
I learned about research funding models, funding sources, funding decisions, as well as about the slow and resistant nature of federal statistical agencies like the Census Bureau, the latency in their results, and growing pressure to innovate (this pressure comes mostly from better statistics being produced by market-driven businesses like Google, Facebook, Twitter etc.)
Good book about why publicly available data should of the highest quality while preserving privacy and what is needed to achieve this. Drawbacks: slightly repetitive and (for non-US citizens/residents) highly US-centered
Starts with strong observations in the introduction about the current state of the federal data and statistics systems, but these same observations are continually repeated throughout the book without much additional insight.