¶ 25. dissenting. The majority decision undermines the ability of the Department to administer the prisons, unnecessarily narrows the trial courts’ traditionally wide discretion in determining factual questions, and inserts this Court into the day-to-day management of the prisons. I respectfully dissent.
*498¶ 26. The principal infirmity in the majority opinion is that it effectively applies a de novo standard of review to the trial court’s factual determination that the Department does not intend to punish inmates by imposing the loaf diet. This is contrary to our law and results in appellate fact-finding.8 Instead, this Court should view the trial court’s factual findings in the light most favorable to the prevailing party, and disregard modifying evidence. N.A.S. Holdings, Inc. v. Pafundi, 169 Vt. 437, 438, 736 A.2d 780, 783 (1999). “Findings are reviewed for clear error, and will not be disturbed even if contradicted by substantial evidence; rather, an appellant must show that there is no credible evidence to support the findings.” Stannard v. Stannard Co., 2003 VT 52, ¶ 8, 175 Vt. 549, 830 A.2d 66 (mem.) (emphasis added). Intent, in both the criminal and civil contexts, is a factual question. See State v. O’Dell, 2007 VT 34, ¶ 11, 181 Vt. 475, 924 A.2d 87; Bixler v. Bullard, 172 Vt. 53, 58, 769 A.2d 690, 694 (2001). By contrast, we affirm legal conclusions if they reflect the correct legal standard and are supported by the findings. Stannard, 2003 VT 52, ¶ 8. Finally, and contrary to the majority opinion, this Court traditionally defers to the construction of statutes by the agency charged with implementing them. State v. Rolfe, 166 Vt. 1, 7, 686 A.2d 949, 954 (1996).
¶ 27. As noted, the trial court found, as a matter of fact, that the Department does not intend to use the Nutraloaf diet as punishment. There was credible evidence before the trial court that supported this finding, and it must therefore be affirmed. Stannard, 2003 VT 52, ¶ 8. Directive 413.09 itself states that its purpose is “to reduce or limit the ability of inmates to misuse bodily waste or food.” There was testimony from both the drafter of the directive and from the Department’s health services director that the directive was not intended to punish, but rather to prevent. The trial court chose to credit this evidence rather than the inmates’ contrary testimony, and this Court should not revisit that choice on appeal. See State v. Dixon, 2008 VT 112, ¶ 34, 185 Vt. 92, 967 A.2d 1114.
¶ 28. The majority’s answer to this is, in part, that the “trial court did not explicitly find that the Nutraloaf regime was without *499any punitive intent.” Ante, ¶ 15. The trial court’s order is to the contrary, however. The court found that “Nutraloaf is . . . not directed as direct punishment,” and that “[t]here is no strong evidence in this case that the Department has intended to use the Nutraloaf diet as punishment.” Nowhere in the trial court opinion is there a finding that the Department has any intent to punish. Finally, the court concluded “that the Department has implemented its Nutraloaf program for legitimate non-punitive purposes.” The majority acknowledges these findings, but dismisses them out of hand as evidence of “mixed motives on the part of the Department with regard to the Nutraloaf program.” Id. As noted, this is directly contrary to the trial court’s explicit factual findings. The majority’s “mixed motives” characterization, even if true, would not compel reversal, however.
¶29. We faced mixed motives, and mixed effects, in State v. Strong, and concluded there that the civil suspension of a driver’s license was not “punishment” because it could “not fairly be characterized . . . only as a deterrent or retribution.” 158 Vt. 56, 59, 62, 605 A.2d 510, 512, 514 (1992) (emphasis added and quotation omitted). We explicitly acknowledged that the suspension functioned, in part, as a deterrent: “Although there is an element of deterrence to the summary suspension of an operator’s license, this element is present in any loss of license or privilege and is not the primary focus of this statutory scheme.” Id. at 61, 605 A.2d at 513. We concluded that summary suspension was rationally related to the “purpose of protecting public safety by quickly removing potentially dangerous drivers from the roads,” that it was not excessive in relation to that purpose, and that “we must defer to the Legislature in determining the remedial action necessary to achieve its goals.” Id. Our holding in Strong resulted from the straightforward application of familiar, common-sense standards concerning “punishment.” Similarly, here, there was no error in the trial court’s finding that there was no intent to punish by imposing the Nutraloaf diet.
¶ 30. The majority is incorrect in concluding that the fact that the diet may be discontinued after three days means that it is punitive. Quite the opposite, as is evident in the analogous context of evaluating civil versus criminal contempt. In that context, the fact that a contempt sanction will end upon the contemnor’s cessation of the offending behavior weighs in favor of the conclusion that the contempt is civil, not criminal, and coercive, not *500punitive. See, e.g., In re Sage, 115 Vt. 516, 517, 66 A.2d 13, 14 (1949). Coercive sanctions may be imposed on the civil contemnor, provided that the contemnor has the present ability to purge himself of contempt by complying with the court’s order. Id. By contrast, criminal contempt imposes sanctions not to compel compliance with an order, but rather to vindicate the dignity of the court, deter future misbehavior, and punish misconduct. Russell v. Armitage, 166 Vt. 392, 407, 697 A.2d 630, 640 (1997) (Morse, J., concurring). Criminal contempt cannot be purged by cessation of the offending behavior. Although the contempt cases do not govern directly, they do highlight the distinction — not made by the majority — between coercion and punishment.
¶ 31. The majority’s de novo scrutiny is nowhere more evident than in its analysis of the “principal aim” of the Nutraloaf program. Despite the trial court’s plain finding that the diet was not intended to punish, and the support in the record for that finding, the majority concludes that “the principal aim of the Nutraloaf program is to deter.” Ante, ¶ 16. This is directly contrary to the trial court’s finding that “[t]he primary goal of the Nutraloaf program is to limit an inmate’s ability to misuse food, utensils, or bodily wastes, by eliminating utensils, [and] by presenting the food in a form that is less messy.” As noted, that finding was supported by the record. The trial court was presented with — and apparently chose to credit — evidence supporting the unsurprising notion that inmates are less likely to throw their own bodily wastes if they have to do so with their bare hands. The majority chooses on appeal to discredit this evidence simply because the program may not be perfectly effective as a preventative measure, and because the Department retains some discretion to provide Nutraloaf in different ways to different inmates based on their particular circumstances. The majority’s logic today is essentially that prevention cannot be the principal purpose of any program that is not 100% effective.9
¶ 32. As we noted in Conway, the next step in the inquiry is whether the government decision is reasonably related to a legitimate government purpose. Conway v. Cumming, 161 Vt. 113, 119, 636 A.2d 735, 738 (1993) (citing Bell v. Wolfish, 441 U.S. 520, *501538-39 (1979)). This inquiry, of course, must take account of the “necessarily broad discretionary authority of prison officials over prison administration.” Id. at 115, 636 A.2d at 736 (citing Jones v. North Carolina Prisoners’ Labor Union, Inc., 433 U.S. 119, 126 (1977)); see also Young v. Coughlin, 866 F.2d 567, 570 (2d Cir. 1989) (“[G]reat deference should be accorded to prison officials as they undertake the difficult responsibility of maintaining order in prisons”). Further, I note that the Commissioner of Corrections is charged by statute with the responsibility “[t]o maintain security, safety and order at the correctional facilities and act to subdue any disorder . . . which may occur at any facility.” 28 V.S.A. § 102(c)(6). The majority, however, does not in any way defer to the Department or acknowledge the discretion that must necessarily inhere for the Department to be able to discharge its weighty responsibility; to the contrary, the majority chooses on appeal to disbelieve the testimony of the very Department officials whom the trial court chose to credit.
¶ 33. The trial court here concluded that “the Department has implemented its Nutraloaf program for legitimate non-punitive purposes.” The court also noted, as we did in Strong, that there are “punitive aspects” to the program, which arise necessarily out of the Department’s need to “avoid creating any incentive for inappropriate behavior.” Just as potential drunk drivers might be deterred by the prospect of summary suspension, inmates might be deterred in some measure from inappropriate food-related behavior by the prospect of being served the loaf diet. But we had no difficulty holding, in Strong, that summary suspension was not punitive. Indeed, we noted with approval the emerging development of a bright-line rule that such summary suspensions, though they may have incidental deterrent effects, have a generally “nonpunitive purpose” that is “clear and compelling.” 158 Vt. at 62, 605 A.2d at 514.10 So it should be here. The nonpunitive purpose, as the trial court explicitly found, is to prevent inmates from having ready access to vessels in which to store their bodily wastes before throwing those wastes at guards or other prisoners.
*502¶ 34. Consistent with our holding in Strong, a federal court in Arizona has twice concluded that loaf diets are not “punishment” and, thus, that no hearing is necessary before they are imposed. Joseph v. Arpaio, 2008 WL 243690, slip op. at *5 (D. Ariz. 2008) (reasoning that because “nutrñoaf’ is not punishment, no process was required); Bugoni v. Coffman, 2006 WL 3333078, slip op. at **8-9 (D. Ariz. 2006) (concluding that pretrial detainee placed on “nutrñoaf’ diet for sixty consecutive days, and several months in the aggregate, was not entitled to hearing because diet was not punitive). Petitioners have directed us to no case, and our research reveals none, holding that a nutritionally complete but arguably unappealing prison diet is punishment.11
¶ 35. Nor does the fact that Nutraloaf is, according to the inmates who may be served it, no more appealing than standard prison fare require us to reverse the trial court’s decision.12 The Department does not bear the burden of ensuring that the food it provides to excrement-throwing prisoners is precisely as appealing as the food it provides to other inmates. Indeed, the only reliable way to ensure that the special management meals would not be less appealing than standard prison food would be to design them to be a great deal more appealing than the usual fare. Even that strategy might founder on the shoals of the simple fact that different people have different tastes in food. The need to avoid such an absurd result is precisely why the cases uniformly recognize that nonpunitive measures may have incidentally deterrent effects without being deemed punitive. See, e.g., Hudson v. United States, 522 U.S. 93, 102 (1997) (sanction need not be “‘solely’ remedial (i.e., entirely nondeterrent)” in order to be deemed nonpunitive); Strong, 158 Vt. at 61, 605 A.2d at 513.
¶36. It appears that any food served pursuant to Directive 413.09 is “punishment” under the majority’s analysis, unless such food is designed to be maximally appetizing, or at the very least *503more appetizing than standard prison food. Any food that is prepared in a manner meant to ensure that inmates do not have an incentive to throw excrement at guards will be deemed punitive. For example, the trial court credited the Department’s evidence that providing sandwiches would provide an incentive for the very misbehavior the program is intended to prevent. The court explicitly found that, “[a]lthough some of the purposes [of the Nutraloaf diet] could be met by serving sandwiches, the court credits the Department’s testimony that the effectiveness of the program would be undermined if the meals were appealing to the point of providing an incentive for misbehavior.”
¶ 37. In sum, the Department has decided to serve inmates who misuse bodily waste a meal that is nutritionally complete, easy to prepare, easy to serve without utensils, and that appears to be economical as well. The majority, via appellate fact-finding, deems this meal “punitive” despite an explicit trial court finding — amply supported by the record — to the contrary. Thus the Department faces a Hobson’s choice of either providing misbehaving inmates with their choice of foods that are likely more appetizing than standard prison fare, thereby encouraging the very behavior that it needs to prevent, or simply doing nothing.13 I do not believe that the Legislature intended § 851 to have such far-reaching effects.
¶ 38. For that reason, and the others detailed above, I would not reverse the trial court. In doing so, the majority prevents the Department from taking swift, temporary, and entirely unhurtful preventive action against profoundly disruptive and dangerous inmate misbehavior. It is quite clear on this record that punitive deterrence is “not the primary focus” of the regulation, and that the special management meals given to these inmates are, therefore, nonpunitive. Strong, 158 Vt. at 61, 605 A.2d at 513. I would affirm.
¶ 39. I am authorized to state that Justice Burgess joins in this dissent.
To cite just one example, the majority refers to Nutraloaf as a “compost” of several ingredients, “mashed together.” Ante, ¶ 2. Nothing in the record supports this pejorative characterization.
Had this logic been applied in Strong, we would surely have concluded that summary license suspension — which does not remove the drunk driver’s ability to reoffend — was punitive. Indeed, it is difficult to imagine any preventative action that could surmount the barrier the majority erects today.
The majority distinguishes Strong on the puzzling basis that “Strong was a case involving mixed effects rather than mixed motives.” Ante, ¶ 16 n.3. The majority today, however, essentially derives purported “mixed motives” from the mixed effects of the Nutraloaf regime. Given that mode of analysis, the motives/effects distinction is not a sound basis on which to distinguish Strong.
To the extent that the Eighth Amendment cases cited by the majority, ante, ¶ 19, appear to answer the question of whether diets like this one are “punishment,” their answers are purest dicta. Those cases, concerned with the question of whether a diet was “cruel and unusual” and therefore prohibited by the Constitution, simply assumed, for purposes of deciding that larger question, that the diet at issue was punitive in some measure. But the question was not squarely raised in those cases, and the majority’s reliance on them is unavailing.
The Department had not, at the time of trial, ever served Nutraloaf to any of the plaintiffs.
The majority suggests “one rather obvious option — that in response to inmate misconduct, the Department may serve standard prison fare not requiring utensils and trays until guilt is determined at a § 852 hearing after which the Nutraloaf regime may be implemented.” Ante, ¶ 23. The majority does not — presumably because it cannot — define “standard prison fare,” which renders the nature of this option somewhat less than “rather obvious.” Indeed, the record is silent as to what the standard fare is in our prisons.