LSE - Small Logo
LSE - Small Logo

Sierra Williams

May 22nd, 2014

Thomas Piketty’s Capital changed the global discussion about inequality because of its great data – now make it open.

0 comments | 2 shares

Estimated reading time: 5 minutes

Sierra Williams

May 22nd, 2014

Thomas Piketty’s Capital changed the global discussion about inequality because of its great data – now make it open.

0 comments | 2 shares

Estimated reading time: 5 minutes

ulrichatzThe rich data informing Thomas Piketty’s landmark research in Capital in the Twenty-First Century has been instrumental to its success. Ulrich Atz argues it is highly commendable that Piketty has made attempts to share the data files. But none of this data is explicitly open for reuse and fails to be available in machine-readable formats. Without an open licence it is not clear whether reuse is permitted, limiting further engagement and research on inequality.

French economist Thomas Piketty has written a bestseller that reached number one on despite competing with books such as Game of Thrones. Capital in the Twenty-First Century has put a spotlight on the widening gap between the rich and the rest. It is a tremendous achievement to ignite the public discussion around inequality, especially in the US.

The book is so promising, some argue “Nobel-prize worthy”, because it uses historic data such as tax records – contrary to ungrounded economic theory. Professor Piketty shows in great detail how inequality evolved over the past 300 years. He writes “[the] debate about the distribution of wealth has long been based on an abundance of prejudice and a paucity of fact.”


Image credit: RoyBoy (Wikimedia, CC BY-SA) using data initially published as Thomas Piketty and Emmanuel Saez (2003)

This is the great power of data: it undermines purely ideological arguments. As we shape the world’s politics with data, transparency and open access are paramount. Even though the data behind the book is public, it is not open. We would love to see a clear licence, better dissemination and simpler formats to enable everyone to engage in the debate around growing inequality.

What’s in the book?

thomas piketty
Image credit: Sue Gardner (CC BY-SA)

Thomas Piketty’s opus magnum includes three major contributions:

  1. Piketty collected and analysed historic data and shows in great detail how inequality evolved over the past 300 years.
  2. He develops the theoretical ideas of why capitalism has a tendency towards the concentration of wealth and how economies will evolve over time.
  3. As an “utopian” solution he proposes a progressive global tax on capital (=wealth).

Among the raving reviews, the book has, not surprisingly, attracted criticism. However, even free-market economists have praised Piketty’s foray into the data and history of wealth, whereas the most ardent critics, e.g. a fund manager, simply ignore his “blizzard of data”.

At its core, Piketty offers a compelling explanation of the, what MIT professor Robert Solow calls, “rich-get-richer dynamic”. He attacks the political right and financial elites with no hesitation. Surprisingly, they have no real answer to his claims that the current trends towards more inequality are detrimental. The data gives Piketty an almost undisputed case, but others need to be able to use it to, to perform their own analyses and double-check Piketty’s claims.

Where is the data?

It is highly commendable that Thomas Piketty publishes the underlying Excel files and also the original sources on his academic site. If you are search-savvy, you can find the data behind the book in a more accessible format on Quandl, a data catalogue.

The main source for income data is a dedicated and growing website, the World Top Incomes Database curated by Piketty and three colleagues. There is a similar dataset at the Luxembourg Income Study Database (LIS).

Versions of the World Top Incomes Database, such as a full dump on, can be found on third-party sites. However, we recommend using the original source because it has the most up-to-date data and acknowledges the work that goes into collecting the data.

What’s the problem?

None of this data is explicitly open for everyone to reuse, clearly licenced and in machine-readable formats.

The data on Piketty’s site is buried in the French section only, and full of cumbersome file formats. There is no licensing information on this site to indicate how the data can be reused, nor on the Quandl copy.

The results of a major data collection exercise, the rights to the World Top Incomes Database are clear in the copyright statement “© The World Top Incomes Database.” In the absence of an open licence it cannot be used without explicit consent. The availability of downloads and the described terms of attribution implicitly encourage re-use, but without a licence it’s not clear whether this is permitted only for research purposes or more widely.

The website and data formats are optimised for humans, with Excel files and point-and-click data selection menus. This is very useful for a quick overview or a specific analysis; a broader use case, however, is virtually impossible without machine-readable data.

In the Luxembourg Income Study Database (LIS), bar for some reports, you only get access to the data if you have a researcher’s account. Moreover, the LIS Microdata User Agreement states, among others, the following terms:

to use the microdata only for scholarly, research, or educational purposes

never to use the microdata for commercial purposes;

The data are not open for everyone. The Upshot, the New York Times’ new data journalism site, published an interactive version of different income distributions. To make this happen, they had to directly collaborate with the LIS team.

What next, Mr Piketty?

We have a few suggestions that would strengthen the book’s cause.

  1. Make the data from your book and in the World Top Incomes Database open using a Creative Commons attribution licence to give users more confidence in how they can use the data
  2. Provide a full data dump for the World Top Incomes Database
  3. Create a set of interactive web graphics that are more engaging than those in the book. Or let others “start touching up the charts like models in Vogue” and feature them
  4. Publish more details on the prediction models and how to replicate them or create your own scenarios
  5. Share the data, metadata and academic sources in machine-readable formats
  6. Go through the process of creating an Open Data Certificate

We believe, then, the success of Capital in the Twenty-First Century will doubtless continue. It will inspire others to contribute to the data collection, allow more people to see the facts about inequality and give those in political spheres the arguments for change.

This piece originally appeared on the Open Data Institute blog and is reposted with the author’s permission.

Note: This article gives the views of the authors, and not the position of the Impact of Social Science blog, nor of the London School of Economics. Please review our Comments Policy if you have any concerns on posting a comment below.

About the Author

Ulrich Atz is the Head of Statistics at the Open Data Institute (ODI) in London. At the ODI he does research, consults startups and governments on the business case of open data, and holds training courses for statisticians, journalists and anyone interested in data. He holds a Diplom in Economics from the University of Mannheim and a MSc in Social Research Methods from the London School of Economics.

Print Friendly, PDF & Email

About the author

Sierra Williams

Posted In: Academic communication | Impact | LSE comment | Open Access

Leave a Reply

Your email address will not be published. Required fields are marked *