A little over a year before the 2027 elections, Javier Milei's government faces the dilemma that sooner or later defines every reformist administration: choosing between what is right and what is convenient.
This is not a choice between good and evil, which would be easy. It is a choice between two paths that, for any government, both feel reasonable. One pays off in the medium term and is costly now. The other pays off now and is paid later, when the one who pays is someone else.
The micro chorus
The Argentine public conversation has revolved around a single word for months: the micro. Public opinion, the political elite, and even PRO partners hammer daily on the same diagnosis. Formal employment is being destroyed in urban areas. Activity is moving in a serrated manner. And the conclusion is always the same: people are getting tired.
It is worth starting by acknowledging that the data is real, because denying it would be exactly the mistake this space criticizes in Kirchnerism. Registered private salaried employment fell to 6.130 million in April, 128,600 jobs less than a year earlier. The irregularity of credit to families reached 12.8% in May, 8.3 points above the same month in 2025. Mass consumption recorded its sixth consecutive year-on-year decline in June. None of this admits relativization.
Many times the key is not to understand what is said, but what is omitted.
What the chorus omits
Let's start with delinquency, which is the data that is most agitated and the worst understood. The irregularity of credit to families exploded because credit to families has returned to existence. In 2023, delinquency was low for the simplest reason in the world: no one was lending. As Minister Caputo argues, the increase in delinquency is a result of banks acting like banks again. The phenomenon is real and painful for households that ended up on the wrong side, but it is the entry price of an economy that is recovering financial intermediation after twenty years without it. The banking sector itself estimates that irregularity will decrease in the last six months of the year.
With employment, we need to dig deeper. The decline in registered private employment did not start with this government: the SIPA report itself indicates that it has been on a downward trajectory since September 2023, three months before Milei took office. Argentina has not created formal private employment for over a decade, under any political sign, and the reason is not cyclical: the regulatory framework for registered employment is a disaster. Hiring under a dependency relationship means assuming a contingent liability of unknown dimensions, and the system systematically punishes the employer who complies.
Interestingly, workers know this and act accordingly. While registered salaried employment fell, the monotributo added 49,500 new registrants in twelve months. Many of them are not day laborers: they are professionals, technicians, and service providers who choose to invoice rather than enter into a traditional dependency relationship. The country stigmatizes "informal employment" as if it were always an imposed misfortune, when in many cases it is a decision.
And the numbers confirm this brutally. According to the INDEC salary index, in May, incomes accumulated a rise of 12.3% in registered private sector, 13.7% in the public sector, and 23.9% in the unregistered private sector since December. Against an accumulated inflation of 14.7% in the same period, the registered worker lost and the unregistered gained by ten points. If informality were pure forced precariousness, that relationship would be reversed. Outside the constraints of collective bargaining and the burdens of the formal regime, incomes move faster. The Argentine problem is not that people flee from registered employment; it is that registered employment is so expensive to offer and so unattractive to accept.
Then there is consumption, the most exposed flank, where it is advisable not to cheat in any direction. Mass consumption is bad, and the official data confirms it: the INDEC Supermarket Survey recorded a year-on-year decline of 0.7% in May and a cumulative decline of 2.8% for the year at constant prices. This is not a fabrication of consulting firms.
What is a mistake is to discuss consumption solely in year-on-year variations. In a country whose comparison base changes direction every twelve months —the first half of 2025 was exceptional because it contained all the postponed consumption of 2024—, the year-on-year measure reflects where you are coming from, not where you are. What matters is the level.
And at levels, the picture is mixed, not negative. Car registrations fell 30.3% year-on-year in July, and the headline wrote itself; however, in the last twelve months, 562,170 vehicles were registered, well above the 414,211 units of all of 2024 and the 449,438 of 2023. The shelves, on the other hand, remain below where they were before December 2023.
There is also a reason to suspect that part of that decline is not a decline but a rotation. We measure mass consumption basically as consumption in supermarkets, and supermarkets lost the artificial advantage given to them by the Precios Cuidados: without that regime, they ceased to be the obligatory channel and found themselves competing equally with wholesalers, neighborhood stores, and, above all, with e-commerce, which grew 34.8% in the first half. Additionally, there is something that no statistic captures well: it is perfectly rational to forgo wine, cold cuts, or Saturday barbecue for a few months to save for a motorcycle, a computer, or a car down payment. This appears in the data as collapsing consumption. It is consumption that is being reordered.
And why is the shopping cart being tightened? Because of income, and here we return to the previous point. The registered real salary has accumulated a year-on-year decline of 3.1% in the first five months of the year, while the unregistered salary outpaces inflation by ten points. The difference is not the cycle: it is the regime. And from there follows the conclusion that discomforts the chorus: if household consumption is tied to formal salary and formal salary cannot rise, the solution is not a subsidized credit plan that pushes the shelves for six months. It is to dismantle the structure that prevents it from rising on its own. A cash plan buys a semester. Labor regime reform solves the problem and costs not a peso of surplus.
Regarding other aspects, the fundamentals do not describe a deteriorating economy. June's inflation was 1.9%, breaking below the 2% barrier for the first time in almost a year, and July would be around that same level despite carrying tariffs, transportation, and winter vacations. The EMAE trend-cycle series has accumulated 26 consecutive months of expansion, the longest cycle since 2011. The country risk closed July at 430 basis points, and the specter of 2027 left the table.
And there is something that no one is pointing out: this growth is occurring without the mechanism that historically produced it. In Argentina, the economic cycle has been synchronized with the electoral cycle —growth occurred when an election needed to be won, through active policies, spending, and suppressed tariffs, and the bill was paid the following year—. This time, the economy is growing with a primary surplus of 0.6% of GDP and a financial surplus of 0.1% in the first half, simultaneous disinflation, and without an artificial exchange anchor. The market projects a GDP 3% higher than the average for 2025. Growing without stimulus in Argentina is the anomaly; at other times, it was called a rebound.
Two countries, one microphone
The political elite speaks as if there were already a popular verdict issued. There is not: an average of thirteen consulting firms places Milei ahead of Kicillof in voting intention by space, by candidate, in floors and ceilings, and in a potential runoff, while the Di Tella government confidence index fell 6.5% in July. There is dispersion, which is something different from defeat.
But the data that organizes the entire debate is not in the electoral polls. It is in the consumer confidence index of the same university, and it is geographic.
In July, the national ICC fell to 40.67 points. Broken down by region: the Interior stood at 45.55, the City of Buenos Aires at 39.83, and Greater Buenos Aires at 38.12. The monthly decline was 6.35% in GBA and 6.24% in CABA, against just 1.68% in the Interior. Seven and a half points separate the Interior from the Conurbano.
A record of twenty-five years, moreover, and signed by a university that no one accuses of being pro-government: the Research Center in Finance documented in May that the gap between the Interior and the Metropolitan Area reached its highest level since the series began in April 2001, attributing it to the energy expansion around Vaca Muerta and the mining boom. What this indicates is uncomfortable for the chorus: there are two Argentine economies coexisting, one that grows based on resources and exports and another that continues to wait. The problem is that the public conversation takes place almost entirely in the latter. It is written, filmed, and edited in the area where the index is worse. When a panel from Buenos Aires says "people are getting tired," it is accurately describing the people around them. The mistake is not the diagnosis; it is confusing the neighborhood with the country.
And from there follows what truly matters. If the discontent is substantially more limited than what the dominant diagnosis suggests, then the political cost of maintaining the course is also limited. The pressure to change it relies on a picture that is not that of the entire country.
What is right and the temptation
In that context, doing what is right has a precise and rather boring definition: to continue. To continue the reformist agenda, to keep lowering inflation, to maintain stability, and to wait —patiently, which is the difficult part— for those conditions to mobilize employment and activity.
Boring, but not arbitrary. The order of the factors is non-negotiable: formal private employment does not recover before stability; it recovers afterward. No serious stabilization process generated employment during the adjustment; it generated it when the adjustment was already done and investment could be calculated over ten years instead of ten days. Advancing that sequence through administrative means does not accelerate anything: it destroys the credibility capital that makes the next stage possible.
The alternative is always available and has an old name. Directed cheap credit, bailouts, cash plans, subsidies, industrial policy, public works as a campaign tool. In other words: replacing the invisible hand of the market with the dirty hand of the State when it comes to choosing winners and allocating resources.
The temptation is not an Argentine trait; it appears in all governments around the world when an election approaches. What is Argentine is that it is burned into the DNA of politics here, to the point that it bu
