New Lex Machina Feature Corrects Erroneous Lawyer Listings in PACER


Lex Machina CEO Josh Becker and GC Owen Byrd at Legaltech in New York last week.

Legal analytics company Lex Machina — which LexisNexis acquired in November — is introducing a new feature today that is designed to correct and improve upon the information PACER provides about attorneys of record in federal district court lawsuits.

Called the Attorney Data Engine, it leverages various data sources extracted from PACER to correct the data that PACER does provide about attorneys.

Owen Byrd, Lex Machina’s general counsel and chief evangelist, told me in a recent conversation that PACER is often inaccurate and sometimes flat wrong in its listings of lead counsel.

Lex Machina analyzed IP dockets in New Jersey and Delaware and found that they failed to include 46 percent of attorneys who have worked on those cases, Byrd said.

The reasons for this vary, he said. PACER is replete with misspelled names of attorneys and law firms. Many attorneys are incorrectly associated with former law firms and many clients are incorrectly associated with their former attorneys’ subsequent law firms. Attorneys admitted pro hac vice are sometimes not listed at all.

To correct these errors, the Attorney Data Engine uses three tools:

  • Signature block analyzer. Every legal filing includes a signature block showing the names and firms of the attorneys. Lex Machina extracts the information from these signature blocks to find the names of all attorneys associated with the filing and their correct law firms. This will initially be rolled out for the districts of New Jersey and Delaware, with other districts to follow.
  • Pro hac vice extractor. When attorneys appear in a case pro hac vice, they are often omitted from PACER’s counsel data, Byrd said. This is sometimes true even when the pro hac vice lawyer is the lead counsel. Lex Machina finds pro hac vice motions in PACER and extracts the information about the attorney.
  • Historical snapshots. When an attorney moves to a new firm, PACER attributes that attorney’s work to the new firm, even when it was actually performed at the prior firm, Byrd said. Lex Machina corrects this by storing daily snapshots of PACER data and analyzing them to determine which attorneys and law firms worked for which clients on which cases at any given time.

Another feature of the Attorney Data Engine is name normalization. Lex Machina’s analysis of PACER data found frequent misspellings of lawyer and law firm names. The law firm Quinn Emmanuel had the most misspelled name of any firm, with 99 variations, Byrd told me. Lex Machina finds these variations and normalizes them so that all listings of a name are found. It also shows you all the variations.

Lex Machina says it spent several years developing this technology and it believes it is the only company that is trying to fix inaccurate PACER data.