On August 26, Argentina’s Chamber of Deputies approved a sweeping rewrite of the BCRA’s charter, 144 votes to 102, with 9 abstentions. The bill now moves to the Senate. It would bar the central bank from financing the Treasury, narrow its mandate to a single goal, “preserving the value of the currency,” raise the threshold for removing its president to two-thirds of both chambers, and eliminate the non-transferable bonds the Treasury has used for decades to draw down reserves.
President Javier Milei called it foundational to ending inflation. The diagnosis behind the bill is correct: Argentina’s central bank has spent ninety years as a tool of whichever government controls it. But the proposed fix has a problem no drafting can solve. It assumes the obstacle to sound money is the wording of a law. It isn’t.
Argentina has cycled through four fundamentally different monetary regimes since the BCRA’s founding in 1935: the pre-BCRA banking system, the nationalized bank under Perón, the Convertibility Plan’s currency board, and the post-2002 floating regime. Each had its own legal architecture. Each produced inflation outcomes ranging from roughly 1% annual inflation before the BCRA existed to an average of 63% a year since 1945. And none of them survived. Convertibility, the episode Milei himself cites as the low-inflation benchmark, had an explicit legal restriction on using monetary emission to finance the deficit. It was abandoned in 2002 anyway, because, as Milei has put it, the arrangement was inconvenient for whoever was in power.
The current bill proposes something Argentina has already tried: prohibiting deficit monetization by law. That prohibition already exists, and it does so at the highest level. Argentina’s 1994 constitution instructs Congress to “provide for… the defense of the value of the currency.” It is an explicit constitutional mandate, in force for over three decades, and Argentina has posted some of the highest inflation in the world under it without Congress ever enforcing it. If a constitutional command to defend the currency’s value can be ignored, what does a statutory version of the same command add?
Legal prohibitions carry a second, quieter problem: they don’t need to be violated to stop working. They can simply be repealed or loosened by the next Congress, the way Convertibility was abandoned in 2002 and the limit on short-term Treasury advances was loosened in a 2012 reform, one the current bill now tries to reverse. And where the rules weren’t formally changed, governments have found ways around them regardless of party. In 2010, the bank’s president was removed by presidential decree after a bicameral committee, sitting with only three of its five members because the Senate seats hadn’t been filled, issued a favorable opinion by a 2-1 vote. The formal procedure was followed. Its independence was not. In 2017, a government that had campaigned on central bank independence pressured its own, nominally independent BCRA into changing its inflation targets mid-year. Today, under a government that calls itself libertarian, day-to-day monetary policy runs through the Treasury Secretariat, and the central bank functions, in practice, as one of its departments.
This is not a partisan pattern. It shows up under Peronist governments and under governments that explicitly reject Peronism. The common thread is not ideology; it’s that in Argentina, written rules, including the constitution itself, do not reliably bind the government of the day. A charter reform that assumes otherwise is solving the wrong problem.
A missed opportunity: Disperse what’s being captured
There is a second lever, distinct from writing better legal restrictions: reducing the value of what a government stands to gain by capturing the central bank in the first place. If seizing control of monetary policy requires coordinating several independent actors instead of pressuring one, capture becomes more expensive and less attractive, regardless of what the statute says.
One way to illustrate this is to borrow the structural logic of the Federal Reserve, an example useful mainly because it is one of the central bank designs most recognizable, and outline what an equivalent redesign might look like for the BCRA. This is offered as one way to illustrate the principle, not as a specific proposal for Argentina, let alone the only or best way to do it.
Start with a board no single administration can pack. Instead of a small board aligned with whoever appointed it, or a single governor, monetary authority sits with a body of seven full-time members serving staggered, fourteen-year, non-renewable terms, with no more than one seat turning over in a given year. A president serving two consecutive four-year terms could fill, at most, three or four of those seven seats within the window where those appointments still matter politically.
Add a policy committee no single party can dominate. Interest rate and exchange rate decisions rest not with the board alone but with a broader thirteen-member committee: the seven board members plus six additional seats, drawn from academia and independent economists, allocated not by the executive branch but split evenly among the three largest parties in the lower house, two seats each, serving six-year, non-renewable terms. No single party can unilaterally control a majority of those six outside seats.
Build in a clause that blocks manufactured majorities. If the president’s own party also holds the largest bloc in the lower house, a realistic scenario, the seats it would otherwise receive under that formula are automatically reassigned to the fourth-largest party instead. The clause triggers whenever one party’s combined seats would reach a majority on the committee; the next-largest party picks up the “excess.”
And make removal public rather than a late-night decree. Removal is restricted to incapacity, proven misconduct, or a criminal conviction, never disagreement over a policy decision, and it must go through a multiparty congressional committee that issues a written, mandatory opinion including any dissent. This is substantively different from just raising the vote threshold to two-thirds within the same body: a higher bar in the same adjudication mechanism doesn’t change who controls that mechanism. An arbitrary removal can no longer be settled by a summer decree. It can still happen, but it stops being free: it becomes a visible, attributable, and politically costlier act.
It’s worth being realistic about what a design like this buys and what it doesn’t. Dispersal doesn’t make capture illegal or impossible, and it still depends on someone, a Senate, a bicameral committee, actually enforcing the procedure when a government decides not to comply. Its main function against a determined executive isn’t to prevent capture; it’s to make capture harder to execute cleanly. When power sits in one place, capturing it is more rewarding but also more visible: there’s a single pressure point, and if it gives way, everyone can see who gave way. When power is dispersed, the opposite trade-off applies: capture becomes more expensive because it now requires coordinating several actors at once, but that same coordination can occur piecemeal and be harder to observe in real time. It isn’t a free upgrade. It’s trading a cheap, visible form of capture for a more expensive and potentially more diffuse one.
Judged against this standard, the bill now before Argentina’s Senate is a linear improvement: real, but only as durable as the political will not to repeal or ignore it, exactly the quality that failed to hold across four previous regimes. Distributing appointment and removal power is a structural improvement, because it no longer depends on that same political will renewing itself every election cycle. But it remains, at best, a way of raising the price of capture, not eliminating it. The 63% average annual inflation Argentina has recorded since 1945, the same figure the government cites to justify this reform, is not the product of badly written law. It is the product of a country where written rules, including the highest one, do not tie the government’s hands. Dispersing monetary power doesn’t change that underlying fact. It makes ignoring the rules slower and more expensive.
The reform most likely to survive a determined political attack is still the one that removes monetary policy entirely from domestic discretion: dollarization. It doesn’t solve every problem, but it eliminates monetary policy as a variable available for capture, with a degree of credibility that no Argentine charter, and no structural redesign, however well dispersed, has managed to produce in ninety years. Ecuador and El Salvador dollarized more than 25 years ago with institutions that, at the time, were not meaningfully stronger than Argentina’s are today. In Ecuador, dollarization has outlasted every constitution the country has had since; it is easier to amend the constitution than to leave the dollar.
Dispersing monetary power beats doing nothing, and it beats what’s currently on the table in the Senate. But if dollarization is the second-best option for Argentina, structural dispersal of the kind sketched here is, at most, a third-best: worth having if the political will to dollarize isn’t there, not a substitute for it.
