In an adversarial system, two advocates represent the interests of competing parties before a judge or impartial tribunal. The tribunal observes the dispute from above the fray to remain impartial. The process traces back to the medieval mode of trial by combat, and has been adopted by most common law countries. By contrast, the non-adversarial system used in civil law allows the judge to directly investigate and interrogate witnesses before rendering a decision.
In the United States, we employ an adversarial system for trials, but use a non-adversarial system in administrative proceedings, such as a hearing for benefits before Social Security Administration. Under 20 C.F.R. §404.900(b), the review process for administration of Social Security benefits should be conducted in a non-adversarial manner. Although administrative hearings are non-adversarial, a claimant has a statutory right to have an advocate present for adjudicatory proceedings under the Administrative Procedure Act. The Act does not impose an obligation on the government to provide counsel, nor is the right to counsel guaranteed by the U.S. Constitution.
State bar rules often impose a duty that an attorney “advocate zealously” for their client’s interests. While the duty to advocate with zeal is favored in adversarial systems, legal scholars recognize that a representative in a non-adversarial proceeding serves a different purpose. A lawyer in a non-adversarial setting may serve as counselor, advisor, transaction planner, government official, or decision maker. In the unlikely event a state bar were to reprimand an attorney for lack of zeal at a federal hearing, scholars also opine that the Supremacy Clause would prohibit states from restraining an attorney’s practice before a federal tribunal.
Menkel-Meadow, Carrie, Ethics and Professionalism in Non-Adversarial Lawyering, 21 Fla. St. U. L. Rev. 153 (1999)
Raines, Robert, Professional Responsibility and Social Security Representation: The Myth of the State-Bar Bar to Compliance with Federal Rules on Production of Adverse Evidence, 92 Cornell L. Rev. 363 (2007)