Another issue in veterans law exists at the intersection of evidence and civil procedure. A central feature of veterans law is its size. When a system handles millions of claims for benefits each year, it is inevitable that a great number of them look very similar. For example, hundreds of thousands of hearing loss claims are filed each year. Is there a better way to handle these than adjudicating them one by one?
The obvious answer from general law is the class action. Michael D. Sant’Ambrogio and Adam S. Zimmerman have taken a shot at making it work in The Agency Class Action, 112 Colum. L. Rev. 1992 (2012). However, veterans law has many features that resist such solutions. The most important is the extraordinarily weak finality. If a class action is lost, who can come back and reopen the claim with new and material evidence? If it is any member, then it is likely that large class actions will never end (unless granted) because there will always be at least one member who wants to try again. What about all those with similar fact patterns who are not part of the class? Perhaps their disability had not yet manifested or perhaps they had not even retired from service yet. Are they bound by the class decision? If so, can they independently pursue reopening? The Supreme Court seems to be discouraging overly large classes of late, so what about much smaller classes? At some point, countless thousands of small classes become more of an administrative headache than just adjudicating each claim separately.
The real problem seems to be lack of global fact finding. Is X as least as likely as not the cause Y if the veteran has a particular symptom history and risk factors A, B, and C? To the extent that you have a clear answer to such questions, every case presenting that fact pattern should have the same answer. Unfortunately, in practice, medicine is as much art as it is science, and when claims are based upon the opinions of individual doctors, veterans with essentially identical conditions can face different results when their claims are adjudicated. For hard questions of medicine, VA should have a speedy and effective mechanism to generate one clear answer that is applied to all similar cases.
One such mechanism is to create presumptions by regulation. Those can ensure that all similar claims receive similar treatment. However, drafting and issuing a regulation is a difficult, multi-year process and can draw a lot of political heat. VA does not have a great track record in drafting timely regulations. For example, it commenced a comprehensive rewrite of its adjudication regulations a decade ago that is still far from complete. When it takes years to draft regulations, how often will the science have changed before the final rule is issued? Even if you produce a final rule, how long will it take to revise it in light of new evidence? The problem seems to be only growing worse, as the time it takes to issue regulations increases and the speed at which new scientific discoveries are announced accelerates.
The idea that most intrigues me is that of having VA simply issue global factual finding through the Federal Register. This approach was endorsed by the Supreme Court in Heckler v. Campbell, 461 U.S. 458 (1983), but seems underutilized. VA obtains over a million expert medical opinions each year. Would not justice be better served if the time of those experts was spent working on published decisions matrices? Image a system that focused initially on collecting basic information that could be plugged into a decision tree that produced results such as, “A Veteran who was exposed to X, will be granted service connection for Y if it manifests within Z years, unless the veteran has risk factors A, B, and/or C.” To resolve the most difficult cases, the decision matrix could even include categories for which the adjudicator was directed to seek an opinion when there were enough conflicting factors to make one prudent. There may be many good reasons why the Heckler approach has not become more common, and further thoughts on its limitations or applicability are welcome.
In any event, the core problem should have a solution: Instead of providing uniform answers to similarly situated veterans, we obtain millions of medical opinions that create a risk of conflicting outcomes. (Whether this is a problem in reality is a great empirical question. I am sure anecdotes are not hard to find, but, as the saying goes, the plural of “anecdote” is not “data.”) This is yet another source of great articles waiting to be written.