From Judge Niemeyer in United States v. Carter, some discussion on relying on empirical data to support the calculation of social cost:

To discharge its burden of establishing a reasonable fit between the important goal of reducing gun violence and the prohibition in § 922(g)(3), the government may not rely upon mere “anecdote and supposition.” United States v. Playboy Entertainment Group, Inc., 529 U.S. 803, 822 (2000). Indeed, in Chester, we remanded the case to the district court for fur- ther development of the record because the government attempted to justify the permanent disarmament of domestic violence misdemeanants with unsupported intuitions rather than tangible evidence. See Chester, 628 F.3d at 683. None- theless, the Constitution does not mandate a specific method by which the government must satisfy its burden under heightened judicial scrutiny. See, e.g., Satellite Broadcasting & Commc’ns Ass’n v. FCC, 275 F.3d 337, 355, 360 (4th Cir. 2001) (requiring minimal empirical evidence where interme- diate scrutiny applied and Congress’ justifications were “both familiar and plausible”). On the contrary, the nature and quan- tity of any showing required by the government “to satisfy heightened judicial scrutiny of legislative judgments will vary up or down with the novelty and plausibility of the justifica- tion raised.” Nixon v. Shrink Mo. Gov’t PAC, 528 U.S. 377, 391 (2000). Even when applying strict scrutiny—requiring a more taxing proof threshold than the one we apply here—the government may, in appropriate circumstances, carry its bur- den by relying “solely on history, consensus, and ‘simple common sense.’” Fla. Bar v. Went For It, Inc., 515 U.S. 618, 628 (1995) (quoting Burson v. Freeman, 504 U.S. 191, 211 (1992)). Thus, while the government must carry its burden to establish the fit between a regulation and a governmental interest, it may resort to a wide range of sources, such as leg- islative text and history, empirical evidence, case law, and common sense, as circumstances and context require. See Staten, ___ F.3d at ___, 2011 WL 6016976, at *5, *11.

In developing its record in this case, the government has chosen not to rely on academic research or other empirical data to demonstrate the connection between drug use and gun violence, even though such evidence is abundantly available. See Yancey, 621 F.3d at 686. To be sure, the record need not be as fulsome as that necessary to justify § 922(g)(9), which was the subject of Chester, because the statutory text of § 922(g)(3) contains an important limiting principle that is absent from § 922(g)(9), as well as from many of the other § 922(g) provisions. Section 922(g)(9) permanently disarms anyone convicted of a misdemeanor crime of domestic vio- lence, even if the defendant has only one remote conviction. Although we ultimately upheld § 922(g)(9) as constitutional in Staten, Chester understandably required the government to make a heightened evidentiary showing before upholding the measure. By contrast, § 922(g)(3) does not permanently dis- arm all persons who, at any point in their lives, were unlawful drug users or addicts. Instead, it only applies to persons who are currently unlawful users or addicts.

…

Nonetheless, the government still bears the burden of showing that §922(g)(3)’s limited imposition on Second Amendment rights proportionately advances the goal of pre- venting gun violence. And we conclude that in this case, the record it made is insufficient. Without pointing to any study, empirical data, or legislative findings, it merely argued to the district court that the fit was a matter of common sense. In view of our decisions in Chester and Staten, we therefore remand this issue to the district court to allow the government to develop a record sufficient to justify its argument that drug users and addicts possessing firearms are sufficiently danger- ous to require disarming them.

This burden should not be difficult to satisfy in this case, as the government has already asserted in argument several risks of danger from mixing drugs and guns. For example, it claimed that due to the illegal nature of their activities, drug users and addicts would be more likely than other citizens to have hostile run-ins with law enforcement officers, which would threaten the safety of the law enforcement officers when guns are involved. It claimed that because drug users and addicts would “necessarily interact with a criminal ele- ment when obtaining their drugs,” their transactions in the black market would present far greater risks of violence (including gun violence) than lawful commerce. While the government did not specifically list the risks, it might be able to show, as found by other courts, that the risks arise from drug dealers seeking to maintain distribution territories and networks, protecting their drugs from theft, enforcing pay- ment, and protecting themselves in a market with unrestrained participants. The government also claimed that the inflated price of illegal drugs on the black market could drive many addicts into financial desperation, with the common result that the addict would be “forced to obtain the wherewithal with which to purchase drugs through criminal acts either against the person or property of another or through acts of vice such as prostitution or sale of narcotics.” Finally, it observed that users of illicit drugs “impair their mental function . . . and thus subject others (and themselves) to irrational and unpredictable behavior,” arguing that persons who routinely subject them- selves to the erratic and irrational effects of mind-altering drugs cannot be entrusted with the responsible use of fire- arms.