Estimating the Social Cost of Corporate Data Breaches

Lina Alkarmi, Armin Sarabi, Mingyan LiuarXiv 2026 · arXiv

TL;DR

This paper introduces a framework to quantify the "social cost" of corporate data breaches—the financial losses, time wasted, and psychological distress externalized onto individual consumers. By analyzing 13 years of empirical data, the authors reveal that while corporate cyber-settlements historically underpaid actual consumer harm by orders of magnitude, a "market saturation effect" of stolen data is beginning to narrow this gap.

What Problem Does This Paper Address?

When a major company suffers a data breach, industry reports typically measure the economic impact using corporate metrics: regulatory fines, legal fees, PR crisis management, and stock price dips. This completely ignores the downstream costs shifted onto the individuals whose data was stolen.

Victims of identity theft (IDT) face unrecovered out-of-pocket expenses, hours spent resolving fraudulent accounts, and even physical or emotional distress that manifests as medical bills. The relationship between a corporate breach and individual harm has long been obscured by the difficulty of tracing a specific fraud event back to a specific data leak. This paper bridges that gap, establishing a rigorous method to calculate the true social cost of data breaches and evaluating whether corporate penalties actually match the public burden.

Methods

The authors constructed their framework using two main building blocks:

  1. A Multi-Dimensional Social Cost Model: Instead of looking only at direct financial loss, the model quantifies three categories of harm:
  2. Direct Financial/Professional Costs: Realized out-of-pocket (OOP) losses plus the cost of hiring legal help.
  3. Opportunity Cost of Lost Time: Hours spent resolving the theft multiplied by the national average hourly wage for that year.
  4. Health and Well-being Costs: Monetized healthcare expenses (medical visits, therapy, and medication) resulting from the distress of victimization.
  5. A Breach-to-Victim Conversion Model: To link corporate breaches to individual victimization, the authors used a Wilcoxon signed-rank test to evaluate changes in IDT frequency following a "mega-breach" (defined as exposing $\ge$ 10 million records). They accounted for a "discovery lag"—the delay between data exposure and the victim noticing the fraud.

To project long-term impacts, they applied a time-based log-quadratic regression with a monthly discount factor ($\alpha = 0.8$) to model how stolen data loses its criminal utility over time as passwords change and credit cards expire. Finally, they introduced a Dynamic Saturation Model to de-duplicate record exposure and estimate the number of unique individuals affected.

To handle inconsistencies across survey waves, the authors developed a data harmonization pipeline:

Fig. 1. Workflow for the longitudinal harmonization and processing of the ITS datasets (2008-2021).

Datasets

The study integrates two primary data pipelines:

  • Identity Theft Data: Six waves of the Identity Theft Supplement (ITS) to the National Crime Victimization Survey spanning 2008–2021. This yielded a harmonized, unweighted sample of 41,091 successful identity theft victims, which was scaled using survey weights to represent the broader U.S. population.
  • Data Breach Chronology: The Privacy Rights Clearinghouse (PRC) Data Breach Chronology database. The authors cleaned and heavily augmented this dataset by cross-referencing it with federal healthcare disclosures (HHS) and state-level filings from Maine and New Hampshire to fill in missing record counts.

The workflow for processing and augmenting the breach data is detailed below:

Fig. 2. Workflow for the processing and augmentation of the PRC data breach chronology.

Key Insights and Ideas

  • The 1-2 Month Discovery Window: The Wilcoxon test revealed a statistically significant spike in identity theft cases exactly 1 to 2 months after a mega-breach occurs. By the third month, the immediate post-breach shock dissipates.
  • The Post-2016 Cost Drop: The average social cost per victim plummeted after 2016. The primary driver was the widespread adoption of EMV chip card technology in late 2015, which crushed high-value, in-person counterfeit card fraud and pushed criminals toward online fraud, which banks detect faster and reimburse more easily.
  • The Market Saturation Effect: As billions of records have leaked over the last decade, the criminal market has become saturated. The marginal "yield" or utility of a single leaked record has drastically decreased because the dark web is already flooded with duplicate identifiers.
  • Vulnerable Populations Bear the Highest Cost: While high-income, highly educated demographics have higher odds of being victimized (due to larger digital footprints and credit limits), low-income and minority victims suffer the highest per-victim social costs. For instance, victims with less than a high school education faced an average social cost of nearly \$850, compared to roughly \$350 for those with graduate degrees.

Key Results

The authors applied their model to establish lower-bound (short-term empirical impact) and upper-bound (long-term projected impact) social costs for three landmark mega-breaches, comparing them directly to corporate settlements:

  • 2009 Heartland Payment Systems (130M records):
  • Settlement: \$107 Million
  • Social Cost: Lower Bound: \$592.7M** | Upper Bound: **\$179.1B (1,673x the settlement).
  • 2013 Target Breach (40M records):
  • Settlement: \$18.5 Million
  • Social Cost: Lower Bound: \$305.8M** | Upper Bound: **\$4.06B (219x the settlement).
  • 2017 Equifax Breach (147M records):
  • Settlement Cap: \$700 Million
  • Social Cost: Lower Bound: \$263.8M** | Upper Bound: **\$1.72B.

The Equifax case highlights the saturation effect: despite exposing more records than Heartland, its upper-bound social cost was two orders of magnitude lower, showing how the narrowing gap reflects a maturing, saturated stolen-data market.

Limitations and Future Work

A primary limitation is that the ITS data relies on self-reported surveys, which are prone to recall bias; notably, 74% of surveyed victims did not know how their information was originally compromised. Structurally, the model treats all compromised records as equal units of risk in its case studies, failing to fully weight the fact that Social Security Numbers (SSNs) carry a much higher long-term risk than revocable credit card numbers (though supplementary analysis showed SSN exposure more than doubles per-victim costs).

Future work includes refining the uniform decay factor ($\alpha = 0.8$) to vary based on the sensitivity of the leaked identifiers and exploring the impact of generative AI on deepfake-driven identity theft.

Summary updated

Abstract

While the size of a data breach is typically measured by the number of (consumer, customer, or user) records exposed or compromised, its economic impact is generally measured from the point of view of the corporation suffering the data breach: cost in crisis management, legal fees, drop in stock price, and so on. This study examines whether it is possible to estimate the true cost, or the social cost of a data breach, measured by the impact on its victims and their out of pocket costs. To accomplish this we establish: (1) the estimation of the average direct financial losses of an identity theft (IDT) victim, including the opportunity cost of lost time, and healthcare expenditures associated with distress associated with identity theft; and (2) the estimation of increases in incidents of IDT that can be attributed to a major breach event. Our findings show that the average social cost per victim has declined significantly since 2016. Furthermore, we find that there is indeed a statistically significant increase in the number of IDTs following a mega-breach event when accounting for a discovery lag of 1-2 months post-breach. Applying our model to real-world cases allows us to estimate an upper and lower bound social cost of specific mega-breach events. We find that for the 2009 Heartland and 2013 Target breaches, even the conservative lower bound social cost estimate exceeded settlements by factors of 5 and 18, respectively. In contrast, the 2017 Equifax breach resulted in a lower bound estimate of $263.8 million, falling well within its $700 million settlement cap. While the Equifax upper bound estimate of $1.72 billion in social cost more than doubles this settlement, the narrowing gap between institutional liability and an incident's social cost provides empirical evidence of a market saturation effect that reduces the marginal damage of individual compromised records over time.
BibTeX
@misc{alkarmi2026estimating,
  author = {Lina Alkarmi and Armin Sarabi and Mingyan Liu},
  title = {{Estimating the Social Cost of Corporate Data Breaches}},
  year = {2026},
  eprint = {2603.21270},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2603.21270},
}
Figure