Maryland funder shifts to case-level reporting to spot service gaps, but grantees raised privacy and technical hurdles

Maryland Legal Services Corporation (MLSC) told grantees that it has shifted from periodic aggregate reports to requesting row-level, case-by-case data to better track where legal aid services are delivered and which client groups are underserved. The change, MLSC staff said, aims to give both funder and grantee a clearer picture of case types, county distributions and demographic outcomes.

MLSC representatives said the move is rooted in the corporation’s statutory mission to allocate state funds in a way that is “stable, economical, effective, and statewide.” Speaker 2, representing MLSC, said the prior aggregate approach could “accidentally tell ourselves comforting stories about progress while harm continues,” and that “We need disaggregated road data” to surface uneven outcomes across geography and demographics.

The transition began with outreach in February 2024 and staged pilots before expanding to a full template and data-mapping process. MLSC provided technology grants and staff support to help grantees adapt; staff described four rounds of technology grants and one “dual reporting” year in which grantees submitted both the new row-level exports and the old aggregates while kinks were worked out. By Jan. 28, 2026, presenters said the office had received about 48 case-level Excel files representing thousands of closed cases.

Technically, MLSC collects Excel or CSV exports from each grantee, requires a “data map” that translates the grantee’s internal field names and codes to MLSC’s standardized values, and uses Tableau Cloud’s wildcard union feature to combine files. Speaker 3 said the team publishes a combined Tableau data source and also writes records into an Azure-hosted data warehouse for longer-term storage and repeatable processing. That pipeline supports dashboards that show statewide KPIs and let users drill down to counties, service levels and selected demographics.

Presenters and attendees described practical benefits for grantees as well: once mapping is complete, organizations can export rows from their case management systems and provide standard outputs with less manual reconciliation. MLSC staff said that approach should reduce repeated table-matching problems that arise with aggregate submissions.

Several technical and governance challenges persisted. Staff reported inconsistent field interpretations across grantees that required expanding the project glossary and repeated mapping feedback. Speaker 4 said some grantees with VOCA-funded programs could not provide all requested fields and had to segment or omit certain items. Staff also noted that older or highly customized databases required extra vendor or developer work, which MLSC partially subsidized through grants.

Privacy and re-identification risk were recurring concerns. MLSC presenters said they explicitly rejected PII sent by grantees (birthdates, Social Security numbers, bank accounts) and, for sensitive case types or very small counties where individuals could be identified, they aggregate results to regional levels instead of county-level disclosure. Speaker 3 noted Tableau Cloud and the planned data warehouse are HIPAA‑compliant platforms and that pro‑level security can limit each user to only their permitted view of data.

Speakers also confirmed the dataset contains unique IDs per case but said there is no cross‑agency client linking by default; MLSC would need additional identifying fields (and careful governance) to join records for a single client across organizations.

MLSC staff emphasized the importance of communication and partner support: the team invested time in explainer videos, one-on-one calls, procedure guides and a data error dashboard that flags records failing to map so grantees can correct problems without rechecking entire files. Presenters said the timing—when IOLTA interest and one-time revenue allowed grants for technical upgrades—was crucial to buying the vendor and staff support needed for the transition.

MLSC framed the effort as iterative. Presenters said the first full batch of row-level data was only recently received, and that insights and funder-level decisions (for example, reallocating resources to underserved counties) are a downstream goal once data quality stabilizes. The session closed with resources and contact information for attendees to request templates and follow-up support.

"This is particularly in the context of comparing outcomes data with things like demographic data," Speaker 2 said, summarizing the clarification that row-level data enables more precise equity analyses. "We need disaggregated road data, because without it, we can accidentally tell ourselves comforting stories about progress while harm continues."