During his two years as a federal appeals court judge, John G. Roberts Jr. has compiled a voting record that places him to the right of the average member of the U.S. circuit courts, according to the first independent statistical study of his voting record.

Roberts was most conservative on issues involving civil rights and civil liberties, the study found. At the same time, President Bush's nominee for chief justice of the United States showed a liberal streak on economic regulation and labor issues; in those matters, he was significantly to the left of the average, according to the study.

"An overall comparison of Judge Roberts' decision-making indicates that he is somewhat more conservative than the average appellate court judge, though not dramatically so," writes the study's author, Kenneth L. Manning, a professor of political science at the University of Massachusetts's Dartmouth campus.

Manning presented his paper Sept. 1 at the 2005 annual meeting of the American Political Science Association in the District. He has since posted it on the Web at www.umassd.edu/cas/polisci/roberts.pdf.

The professor studied Roberts's votes in 145 of the 190 cases he has heard at the U.S. Court of Appeals for the D.C. Circuit, excluding 45 that "did not reveal a clear liberal/conservative dimension."

He found that, where such an identifiable ideological dimension did exist, Roberts made a conservative decision 67.1 percent of the time, or about 30 percent more frequently than the average federal appeals court judge. Manning based the comparison on data on the votes of all appeals court judges that have been gathered by the Inter-University Consortium for Political and Social Research at the University of Michigan.

Employing what he said were conventional political science categories, Manning classified all 145 of Roberts's cases under three headings: There were 13 "civil rights/civil liberties" cases, 44 "criminal justice" cases, and 88 "economic activity and labor regulation" cases.

In civil rights and civil liberties cases, as in the criminal justice area, Manning counted a vote in favor of the government as "conservative."

Roberts voted for the government 86.4 percent of the time in criminal cases and 84.6 percent of the time in civil rights and civil liberties cases.

Roberts's score on criminal justice cases was in line with the federal appeals courts generally. But his civil rights and civil liberties votes make him almost four times as conservative as the average appellate judge on such issues, Manning said in an interview.

He cautioned that this score was based on a small sample, because few such cases reached the D.C. Circuit since Roberts joined it in 2003.

By contrast, Roberts's relatively liberal score on economic activity and labor regulation reflected his votes in 88 cases, a large number indicative of the D.C. Circuit's prominent role in reviewing decisions of agencies such as the National Labor Relations Board and Federal Communications Commission.

In those cases, Manning said, Roberts tended to side with government out of the same deference to agency decisions that prevails generally on the court.

"The general read I got was of a non-activist stance -- a general reluctance to go out of his way and rule against government regulators," Manning said. "If the EPA ruled against the chemical industry, the general tendency was to defer to the agency."