When the VA misrepresents performance, veterans suffer

Administrative judges appear to be performing better than ever, in spite of huge increases in performance quotas. The Trump Justice Department implemented an annual “production quota” of 700 cases for immigration judges. The Social Security Administration requested its disability judges to increase their output by almost 20 percent over the past few years. And the Board of Veterans’ Appeals (BVA) increased its output by a whopping 62 percent in one year, deciding 52,661 cases in 2017 and 85,288 cases in 2018.

The Constitution demands that agencies decide cases accurately. And in spite of the production increase, the BVA reported an “accuracy rate” of 94 percent for 2018. According to this reported rate, only 6 percent of BVA decisions contain legal mistakes. If true, these statistics would mean that agencies deciding hundreds of thousands of cases each year — more than all federal courts combined — can provide high quality justice for immigrants, veterans and the disabled at unprecedented rates.

{mosads}Has the BVA finally cracked the constitutional code of mass adjudication? Nothing could be further from the truth. Our research teams at Stanford and UCLA unearthed data on nearly 600,000 cases never before studied by outside researchers, as well as hundreds of pages of agency documents. Drawing on this information and in-depth interviews with agency officials, our research shows that the BVA is seriously misrepresenting its performance.

Veterans suffer from this misrepresentation. The volume of veterans’ appeals is huge. The vast majority are related to disability compensation claims. Some 90 BVA judges decide cases, with an “inventory” of over 425,000 cases pending. A veteran must wait seven years for the BVA to decide his or her case. If the decision is incorrect, the veteran must hire a lawyer, appeal yet again, and spend years to get the error fixed. Veterans are caught in this “churn” of appeals, creating extreme cynicism among some veterans groups. In one veteran’s assessment, VA’s error-prone case handling consists of “delay, deny, hope they die.”

Our research is the first to rigorously examine the BVA’s “quality assurance” program, which generates its accuracy statistics. We worked directly with the former chief of the BVA’s Office of Quality Review to test its efficacy. The program randomly sampled 5 percent of decisions for an elite team of attorneys to review. Attorneys looked for errors, namely judicial decisions that were reversed or remanded by the Court of Appeals for Veterans Claims (CAVC), the court that independently reviews the BVA’s decisions. When they found errors, the attorneys would write memoranda to suggest corrections.

Our results are sobering.

First, the quality review program was ineffective in improving decisions. A decision that underwent quality review was just as likely to be appealed and reversed by CAVC as any other decision. Even if a quality review attorney pronounced a decision error-free, that decision was about as likely to be appealed as one that was not reviewed.

Second, the quality review team graded opinions on a very forgiving curve. The BVA program identified defective rationales in decisions at one-sixth the rate as CAVC did for the same decisions. And when veterans appealed cases that the program deemed error-free, CAVC still reversed and remanded three-quarters of the time.

Third, as production expectations increased, morale plummeted. A typical case for veterans benefits involves thousands of pages of exhibits and numerous complicated legal issues. A veterans law judge must decide 25-30 of these cases each week. Commenting on the impact of rushed decision-making on the quality of his work, one veterans law judge confided that he found himself forced to sign decisions he never would have signed in earlier years: “I could have integrity here or I could stay employed.”

Why would an agency administer a program to such little effect? The answer is as obvious as it is troubling. The BVA’s rosy accuracy rate of 91-95 percent figures prominently in its public relations efforts. The BVA has emphasized this rate in annual reports and congressional oversight hearings. The program has kept Congress at bay with the appearance of effectiveness.

The tragedy is that the program gave the BVA no incentive to improve and Congress no meaningful performance measure to evaluate. If the board misrepresents its performance, other agencies may do so as well. Veterans, immigrants, disabled workers — and all Americans — deserve better. Agencies should be required to review the quality of their decisions and adjudication in the face of systemic problems.  

Agencies need to disclose enough information so that judges, policymakers and the public can verify that these programs promote what the Constitution demands.

Daniel E. Ho is the William Benjamin Scott and Luna M. Scott Professor of Law, professor (by courtesy) of political science, and senior fellow at Stanford Institute for Economic Policy Research at Stanford University.

David Marcus is a professor of Law at UCLA School of Law, specializing in civil procedure and administrative law.