Sapin II: a very opaque transparency bill in France

Data-driven stories outlined how the reporting voted in Parliament could have allowed companies such as Total to hide a large part of their activities

Public reporting requirements in France and the opportunity for a stronger transparency bill

Between 2012 and 2014, France was considered a champion of corporate transparency by its European peers. In 2013, the French Parliament passed a law setting a public country-by-country reporting (CBCR) for banks which likely influenced the vote for the EU Capital Requirements Directives (CRD) a few months later. Public CBCR is widely seen as an efficient way to monitor tax strategies of multinational companies, forcing them to disclose information on their operations (turnover, profits, taxes, subsidiaries, etc) on a country-by-country breakdown (including tax havens). In 2014, France was one of the first countries to transpose the EU Transparency and Accounting Directives, setting up public reporting for extractive companies, different from CBCR but equally useful, commonly known as Payment to Government reporting (PtG reporting).

However, despite promises from various government members and President Hollande himself, the extension of public CBCR to cover all large multinationals was never voted in Parliament. In mid-2016, in the aftermath of the Panama Papers, the French government introduced a new bill on transparency, the anti-corruption fight and the modernisation of economic life, commonly known as the “Sapin II bill”. Initially reluctant to establish a public CBCR, the government eventually backed down and introduced a watered down version of the measure.

An amendment watered down by exemptions

Although fairly technical, the concept of public CBCR has increasingly become a very public transparency issue. Whilst NGOs have demonstrate to the public opinion how public CBCR can increase monitoring and accountability of multinational companies, peakr, or industry, associations have regularly flagged CBCR as a threat to competitiveness. On the eve of the plenary vote, the newspaper 20 minutes published an article called “Tax Evasion: NGOs v Multinationals who will win over the Parliament?”

By introducing a watered down version of public CBCR, government claimed it was meeting everyone’s expectations halfway. But our data-stories illustrate how the provision introduced by the government would have allowed large companies to hide a considerable part of their activities.

Data-driven stories: how the reporting would have allowed Total to hide a large part of its activities

The government’s exoneration included a safe harbour provision. Thanks to this, companies reporting less than a certain number of subsidiaries in a given country (two, three, four, etc., as set by decree) would have been exonerated from disaggregating their report at country level, on the basis that it could represent a threat to their competitiveness.

The Data Extractor created a data-driven story to build the evidence outlining how this exoneration would make the reporting essentially useless. Using data published by the company itself, we outlined that, using the lowest threshold possible (less than two subsidiaries), one of the largest French oil companies, Total, would have been exonerated to report in more than 30 countries, one third of the countries where the company operates.

How did we build the case study? We used a wide range of programs that we learnt through the Data Extractors program:
1. Scraping the list of Total Subsidiaries from their annual report: Scraping data allows to turn data in rigid format (PDF, HTML) into open data. For PDF data, Tabula is usually the most performing program, but does not work all the time. Free online alternatives exist, such as OCR tools. In our case, Tabula did not work and we had to resort to online OCR tools.
2. Cleaning the dataset: it’s often necessary with online scraping tools that are less efficient than Tabula. We used OpenRefine, which also allowed us to organise the datasets and add some filters to the 900 entries.
3. Visualising the dataset: we used Tableau, with support from Open Oil, to create an interactive map outlining which countries would be excluded from Total’s reporting.

Interestingly enough, most of the countries where Total would be exonerated from reporting are countries where Total has extractive assets, countries where transparency is needed the most.

Impact of the data-driven story

During the day of the vote, a large number of MPs used our policy note and our figures on Total to demonstrate how inefficient the provision put forward by the government would be. Our position outlined the inadequacy of a reporting that would exonerate the biggest company from disclosing 1/3 of its activities. The figure was also picked up in a number of newspapers. A group of MPs introduced an amendment to set up a full CBCR, without exemptions.

Unfortunately the amendment did not pass. Just before the vote, the government suspended the debates, called a few pro-government MPs and ask them to come back to the Parliament to eventually outvote the amendment by a couple of people (it was actually the second time they resorted to this tactic to counter a vote on public CBCR after December 2015).

Looking forward: lessons learnt and future opportunities

This data-driven case study shows how important it is to provide clear examples of law implementation of a fairly technical issue to our supporters, media, donors, but also, and primarily, MEPs. No one contested our figures that plainly illustrated the risks of granting exonerations. The advocacy carried out by civil society didn’t pay off as the amendment was rejected. Even worse, the French Constitutional Council declared the provision unconstitutional until its application at the EU level. Our advocacy work is now turning to the EU where debates on public CBCR are starting. The EU commission is supporting a version of public CBCR that is fairly similar to the one proposed by the French government.

This data-driven story is an example on how to find and present complex information to influence debates on CBCR.