Meta Settlements and the Absence of a Method

On October 21, a New Mexico judge ordered Meta to pay $25 billion in civil penalties after a jury found that the company had repeatedly misled Facebook users about its privacy protections and other platform practices, in a case that grew out of the Cambridge Analytica scandal. 

The jury identified nearly 44 million violations of the state’s consumer protection law, arising from misleading representations made over an 11-year period. The state sought between $35 billion and $40 billion, but Meta argued that the penalty should be capped at $3.45 billion. Judge Francis J. Mathew found that the statute’s $5,000-per-violation maximum would support more than $219 billion, then set the penalty at $25 billion. The judgment accrues 15 percent annual interest, and Meta has said it will appeal.

The judgment does not explain how the court moved from $219 billion to $25 billion, and that is a remarkable gap in a country that relies so heavily on litigation as restoration. Twenty-five billion dollars is a number of almost incomprehensible magnitude to most of us. It is large enough to become part of the historical record of corporate accountability. We experience such a number almost more emotionally than analytically.

And the process of arriving at that number offers almost no insight into the economic value of the underlying harm. Can the people be restored by $25 billion? $40 billion? What, precisely, is being measured? The legal system has answers to questions about statutory violations, permissible penalties, deterrence, and constitutional limits. But there is a question that remains unanswered. 

What did the underlying conduct actually cost society?

We have developed astonishingly sophisticated systems for calculating the value companies create for their shareholders. We can measure revenue, earnings, margins, assets, and liabilities with a precision that would have been unimaginable a century ago. Vast institutions exist to scrutinize those calculations, and entire markets organize themselves around their results. Yet when we turn our attention to the costs that economic activity imposes beyond the boundaries of the company, that precision begins to disappear. We are left with judgments, settlements, and regulations. Each tells us something, but nothing tells us everything.

A price set by a roll of the dice

Meta’s repeated rounds in court are a particularly vivid illustration of the insufficiency of our calculations and the subjectivity of economic justice. In 2019, the Federal Trade Commission imposed a $5 billion penalty on Facebook for violations of an earlier privacy order. Now, seven years later, the court arrived at $25 billion. What will the settlements look like in another seven years? In 10? Will the calculations be any more accurate?

We are talking about differences of tens of billions of dollars. And while each figure has a legal rationale, none represents a comprehensive calculation of what the underlying conduct actually costs society or who directly benefits from the judgments.

In August, a coalition of 47 states, the District of Columbia, and three territories announced a separate agreement resolving claims involving both children’s safety on Meta’s platforms and the company’s handling of personal information. The settlement guarantees at least $12.1 billion over 10 years, potentially increasing to $17.1 billion if specified conditions are met. It also requires substantial changes to the design and operation of Facebook and Instagram.

That agreement contains the sharpest comparison. It includes a separate payment of about $459 million that resolves Cambridge Analytica privacy claims for 46 states and two territories combined. New Mexico, which pursued its own case over related conduct, was awarded $25 billion — more than 50 times as much, for one state. That is nearly $12,000 for each resident of the state, against less than $2 per resident of the jurisdictions that settled. A 6,000-fold variance reveals the utter absence of a defensible method.

The figures diverge even within New Mexico. Earlier this year, in a separate trial over harm to children, a Santa Fe jury imposed $375 million in civil penalties, about one and a half percent of the privacy award. Texas, meanwhile, settled separately with Meta for $1 billion.

Under New Mexico’s constitution, any money recovered from the $25 billion judgment must go to education; Attorney General Raúl Torrez has said it will fund school construction. Within the August multistate agreement, the allocation of proceeds varies by jurisdiction. Delaware has committed roughly three-quarters of its settlement payments to youth health and safety initiatives, including crisis intervention, after-school programs, and services for vulnerable young people. Massachusetts has identified youth mental health services, digital literacy, and education among its intended uses, while also acknowledging that a portion of its recovery will enter the state’s general fund. Across the agreement, about 70 percent of the money, up to $11.7 billion, is left to the discretion of state officials, and only 23 percent is directed specifically at youth mental health or social media harms.

These are consequential decisions, and none represents an inherently bad outcome.Funding public education, supporting mental health services, enforcing consumer protection laws, and compensating legal counsel can all serve legitimate purposes. But what determines the relationship between those expenditures and the harm that gave rise to the litigation?

A settlement establishes a financial obligation, but the subsequent allocation of its proceeds reflects another set of decisions, shaped by legal requirements, institutional priorities, and public policy. Those decisions are often almost entirely opaque from the outside.

Crucially, none of these processes begin with a systematic accounting of the economic consequences experienced by the people and institutions affected.

And there is another distinction worth making. The multistate agreement requires changes intended to reduce future harm, including restrictions on minors’ use of Meta’s platforms and independent oversight of the new protections.

Those requirements may prove enormously valuable. But their value cannot be inferred from the size of the financial settlement. To understand their economic significance, we would need to estimate the harm they are expected to prevent. But if the states cannot estimate with any confidence the harm these platforms have caused, how can we possibly put a figure to future harm?

All these proceedings involve different claims and remedies, so their financial outcomes cannot be treated as competing estimates of an identical injury. But that infinitely complex subjectivity is very much the point. A list of settlement figures looks almost comically disparate, and speaks not to the difference in harms absorbed by each state but to the variance in legal proceedings. 

This is the peculiar lottery we have constructed around corporate harm. Even when billions of dollars change hands, we still lack a coherent account of what was lost, who bore the cost, and what it would take to repair the damage.

The Sackler pattern

The opioid epidemic offers another illustration, although here the distance between legal resolution and economic consequence is especially difficult to contemplate.

In 2025, Purdue Pharma and the Sackler family reached a settlement framework valued at approximately $7.4 billion to resolve extensive litigation arising from the company’s role in the opioid crisis. The sum is substantial. It is also, when weighed against the economic and social devastation associated with the epidemic, an insult. 

In 2022, the U.S. Congress’s Joint Economic Committee estimated that the opioid epidemic had cost the United States $1.5 trillion in 2020 alone. That estimate encompasses harms attributable to many causes and actors, over a vastly broader population and period than those addressed by Purdue’s settlement. It would be analytically indefensible to attribute the entire amount to Purdue, or to interpret the difference between the figures as an unpaid balance owed by the company.

But the comparison illuminates the extraordinary scale and fundamental inequity of the accounting problem. 

The costs of opioid addiction do not reside neatly within a single lawsuit or corporate balance sheet, because they are distributed across healthcare systems, families, employers, public institutions, and communities. Some of those costs can be estimated with considerable sophistication. Others remain difficult to identify, attribute, or even begin to express in monetary terms.

What emerges from litigation is, therefore, a legal and negotiated allocation of responsibility, constrained by the evidence, the law, the resources of the defendants, and the practical needs of those seeking relief. The courtroom can determine what a defendant must pay. It cannot tell us everything that has been lost, who has borne those losses, or how much of the burden remains elsewhere.

Costs that remain elsewhere do not simply disappear.

The costs we are taught not to see

The language of externalities — of corporate consequences — can make the problem sound abstract, almost bloodless. But the people and institutions absorbing those consequences experience nothing abstract about them. A cost excluded from a corporate transaction may reappear as a hospital bill or a damaged ecosystem. It may surface immediately, or it may accumulate over years before anyone recognizes its source.

And because the cost has been separated from the activity that helped produce it, the people bearing it may have little practical means of identifying its origin, anticipating its magnitude, or ever recovering what they have lost. This is the fundamental asymmetry at the heart of externalized harm.

The benefits of an economic activity are measured and allocated with extraordinary care while the associated costs are dispersed so widely, and over such long periods, that they become difficult even to identify as parts of the same economic event.

Those barely detectable pathways are a direct consequence of how we have chosen to account for economic activity. A company can report a profitable year while some portion (sometimes an enormous portion) of its costs is being absorbed by insurers or hospitals or court systems or communities. 

The transaction looks different depending on which ledger we examine, but the underlying economy has only one set of consequences.

What a method looks like

A credible method for measuring externalized harms must begin by tracing their causal relationships to the activities that produce them and translating those relationships into estimates grounded in published evidence. It means distinguishing between what is known, what can reasonably be inferred, and what remains uncertain. It means making assumptions visible, establishing consistent categories of analysis, and allowing estimates to be examined, challenged, and continuously improved.

Above all, it means producing information that can be used before the consequences have become irreversible, or before litigation has become the only remaining means of allocating their costs.

Consider what that would make possible.

An insurer evaluating an emerging risk could examine the potential economic consequences of an activity before those consequences materialize as claims. An investor could assess exposures and sources of value that conventional financial reporting leaves outside the frame. A government could compare the expected cost of an intervention with the losses it might prevent. A business could begin to understand not merely whether an activity is profitable within its own accounts, but what costs it creates elsewhere, and whether changing its practices might produce economic benefits that are currently invisible.

A cost that can be identified early can be managed differently from one that emerges only after years of damage and litigation. And an economy capable of seeing more of its own consequences is an economy capable of making better decisions about them.

New Mexico’s $25 billion judgment may ultimately prove to be a landmark in corporate accountability. The multistate settlement may produce meaningful protections for children. The opioid settlements may provide resources to communities that have endured enormous suffering. And these outcomes matter, but they do not resolve the deeper accounting problem.

For all our extraordinary advances in financial analysis, we still rely on mechanisms that respond to the broader consequences of corporate activity without objectively measuring them. Litigation addresses impact (at least in theory), but it does so only after the fact, and it is subject to the mercurial nature of human opinion and the atomization of judicial authority. It is retrospective, adversarial, and constrained by the claims a court is empowered to consider. 

The economic consequences of corporate activity begin accumulating long before a court establishes liability — they are absorbed by people and institutions whose losses may never appear in the financial statements of the companies whose activities contributed to them.

We have come to rely on legal decisions as though they could tell us what the underlying harm was worth. But they cannot do that work alone. The absence of a consistent method leaves us arguing over numbers that may be legally meaningful but economically incomplete, while the consequences themselves continue to accumulate.

The question is why we have accepted a system in which the consequences of economic activity can remain largely invisible until someone has the resources, the standing, and the persistence to bring them into a courtroom.

We know how to measure what companies earn. It is time to become equally serious about measuring what their activities cost. And it is time to do that accounting while there is still something we can do with the answer.

At Calx, this is the work we are building toward: A common, sourced accounting of harm that parties can use before a case is filed rather than after a verdict. If your work depends on numbers like these, write to us at hello@calx.info.