How Argentina Wrote Its Deregulation Laws in Microsoft Word

Introduction

When Javier Milei assumed the Argentine presidency on December 10, 2023, observers expected the usual lag between inauguration and legislation. Instead, within ten days, his administration deployed a sweeping "mega-decree" (DNU 70/23) that instantly deregulated vast sectors of the economy.^[4][5]

The decree was not drafted in those ten days. It was the product of a two-year drafting operation run almost entirely through a single tool: Microsoft Word's Track Changes feature.^[5] This article traces the methodology behind what became one of the most abrupt regulatory shocks in modern economic history.

The Epistemic Filter: Only Redlined Text Accepted

The defining feature of the project was a strict, non-negotiable rule imposed by its architect, Federico Sturzenegger: no meetings, no conceptual memos, no PowerPoint presentations, and no bullet points. The only deliverable the team would accept was redlined legal text.^[5]

Sturzenegger described the mandate in his RedNIE working paper and subsequent lectures at Princeton and the Harvard Kennedy School:

"download the law from the government website, put it into a Word document, and using track changes, rewrite it exactly as you propose it should be."

Each specialist was given a sector — energy, telecommunications, agriculture, labor — and told to download the relevant statutes, open them in Word, and use Track Changes to delete or rewrite every clause they believed was extractive or obsolete.^[2] This forced abstract policy desires to be immediately translated into actionable legal code. If an expert believed a regulation was stifling competition, they could not write a diagnostic essay; they had to physically delete the offending clause and replace it with new, market-oriented phrasing.^[1]

Why Microsoft Word?

Word's Track Changes feature was not chosen for its sophistication. It was chosen because it imposes a particular discipline: every proposed change is visible alongside the original text. Deletions appear as strikethroughs, insertions as colored additions. There is no way to hide vague intentions behind abstract language — the original statute and the proposed revision coexist on the same page.^[5]

This transparency served a dual purpose. First, it made every contributor accountable for specific legal changes rather than general recommendations. Second, it allowed the core team to compile, review, and integrate hundreds of disparate sectoral reforms into a unified package, because every contribution arrived in the same format with the same granularity.

The 70% Attrition Rate

When contributors learned that the only acceptable deliverable was a redlined Word document, approximately 70 percent of them vanished from the project.^[5]

Many individuals who were highly critical of Argentina's regulatory status quo in theory were either unwilling or unable to commit to the meticulous work of legally dismantling it line by line. The methodology quickly separated political pundits from legal engineers.

The 30 percent who remained formed a highly effective drafting corps. Over two years, they worked through the national legal corpus, identifying provisions that clearly reflected special interests rather than genuine public goals. In many cases, the regulatory bias was so extreme that the required fix was obvious once basic economic reasoning was applied.^[5]

The "Scrap Pile" vs "Fix Pile" Taxonomy

As the redlined documents flowed back, the core team sorted them into two categories that became the architecture of the entire reform package:^[1][5]

• The "Scrap" Pile — Laws deemed fundamentally flawed, historically obsolete, or purely extractive. These were slated for total repeal. They later became the foundation for the "Ley Hojarasca" (Dead Leaves Law) and parts of DNU 70/23.
• The "Fix" Pile — Laws governing necessary state functions that required heavy modification to introduce market competition and eliminate rent-seeking. These were integrated into the "Ley Bases" (Omnibus Law) with new text already drafted via Track Changes.

The taxonomy reduced thousands of pages of legislation to a binary decision: does this law get scrapped entirely, or does it get redlined into something that works? There was no third category for "study further."

Coordination Without Meetings

The project operated with no meetings, no central office, and no formal coordination infrastructure. The Word files were the coordination mechanism. Each specialist worked independently on their assigned statutes. The core team at UdeSA received, reviewed, and compiled the redlined outputs.^[1][2]

This decentralized structure was not incidental — it was the point. Sturzenegger deliberately avoided the organizational overhead that bogs down traditional policy development: standing meetings, working groups, steering committees, and consensus-building processes.^[5] The format of the deliverable (a redlined Word document) contained all the information the core team needed. There was nothing to discuss that wasn't already visible in the tracked changes.

The project's budget reflected this lean structure. When it launched at UdeSA in late 2021, funding was exactly zero.^[5] Even after expanding to 150 specialists under Patricia Bullrich's presidential campaign in May 2022, the coordination model remained the same: send out a statute, receive a redlined file.^[2]

The UdeSA Genesis: 7 People, Zero Budget

The project began in late 2021 at the Universidad de San Andrés (UdeSA) in Buenos Aires, roughly two years before the December 2023 presidential inauguration. Federico Sturzenegger — MIT PhD, former Central Bank president, professor at UCLA, Di Tella, and Harvard Kennedy School^[3] — assembled a team of seven:^[5]

1. Sturzenegger himself
2. A professor of constitutional law
3. Four undergraduate law students
4. Sturzenegger's teaching assistant from a Principles of Economics course

This group began the foundational work of reviewing the entire legal corpus of the Argentine Republic, cataloguing historical regulations — some dating back decades — and identifying laws that were functionally obsolete but remained on the books, creating friction for businesses and citizens.^[1]

By May 2022, when the project aligned with Patricia Bullrich's presidential campaign, the team expanded to the 150-person network that would produce the final drafts. But the methodology never changed from what those first seven people established: download the law, open it in Word, use Track Changes, return the file. Read our companion article profiling the architects behind the deregulation program →

Implications for Regulatory Reformers

The Argentine experiment demonstrates a counterintuitive finding: constraining the format of policy contribution can dramatically increase its quality and speed.

The Track Changes methodology worked because it solved three problems simultaneously:

• It eliminated vagueness. You cannot propose a vague reform in Track Changes. Every deletion and insertion is specific and visible.
• It standardized contribution. Whether the contributor was a labor economist or an energy lawyer, their output arrived in the same format with the same granularity. This made integration trivial.
• It filtered for commitment. The 70% attrition rate was a feature, not a bug. The methodology self-selected for contributors willing to do the tedious, line-level legal work.

For any reform team — in any jurisdiction — considering how to convert policy intent into deployable legislation, the lesson is structural: require concrete legal text as the unit of contribution, and the rest of the process simplifies. Read how this portable reform package was transferred from Bullrich to Milei →