←Back to Data Stories
General

How Corruption Moves Through Government

Published: 2026-07-07 • Methodology: Longitudinal OLS Regression Models

Figure 1.1: description

Across Sandiganbayan cases from 2016 to 2025, corruption does not always appear as a purely individual act committed by one public official acting alone. In many cases, it looks more like a process that moves through different parts of government. Out of 1,343 cases in the dataset, 45.9% involved more than one accused public official. When the cases are grouped by role family, the pattern becomes even clearer: 691 cases, or 51.5%, involved at least two distinct role families, while 368 cases, or 27.4%, involved at least three. This does not mean that every case is coordinated, system-wide, or equally complex. But it does show that a substantial share of Sandiganbayan cases involve more than one bureaucratic function, suggesting that corruption often passes through institutional channels rather than remaining confined to a single office or official.

The role-family patterns also show where these institutional links tend to form. Elected or political officials appeared in 637 cases, senior agency management in 620 cases, and finance/accounting/treasury officials in 543 cases. The most common co-occurring pair was finance/accounting/treasury with senior agency management, appearing together in 307 cases. This was followed by elected or political officials with finance/accounting/treasury officials, which appeared together in 227 cases. These pairings point to a recurring relationship between political authority, managerial approval, and financial processing. In the cases, alleged irregularities often moved through ordinary administrative steps: signatures, vouchers, certifications, procurement papers, inspections, disbursements, and approvals. The dataset therefore supports a more institutional view of corruption. Accountability still involves individual officials, but the pattern also points to the bureaucratic chains that allow irregular transactions to pass through the normal machinery of government.

Access our dashboard to see the full dataset