concurring in part and dissenting in part:
The majority holds that Bardzik was a policymaker while serving as Reserve Division Commander, but was not a policymaker while serving in Court Operations. I agree that whether Bardzik was a policymaker while serving as Reserve Division Commander is a close question. See Maj. op. at 1146. However, resolving all factual disputes in Bardzik’s favor, I would conclude that he was not a policymaker in either the Reserve Division or in Court Operations. Thus, I dissent from the majority opinion in part.
To conclude that Bardzik was a policymaker while commanding the Reserve Division, the majority analyzes Bardzik’s duties under the factors stated in Fazio v. City of San Francisco, 125 F.3d 1328, 1334 n. 5 (9th Cir.1997) (“vague or broad responsibilities, relative pay, technical competence, power to control others, authority to speak in the name of policymakers, public perception, influence on programs, contact with elected officials, and respon*1153siveness to partisan politics and political leaders”). The majority concedes that some of the factors favor Bardzik: he did not have a relatively high salary, he did not have authority to speak on behalf of policymakers, and he did not enjoy public notoriety. Maj. op. at 1146. The majority also holds that the factor of “responsiveness to partisan politics and political leaders” is inconclusive. Maj. op. at 1146.
In my view, the remaining Fazio factors also fail to support the conclusion that Bardzik was a policymaker with reference to the Reserve Division.
1. Control of Others
The majority argues that Bardzik had the “power to control others” as Reserve Division Commander. Maj. op. at 1148. But in that position, Bardzik supervised only one more fulltime paid employee than he did in Court Operations, a position the majority holds was not policymaking. Maj. op. at 1151. Thus, considering full-time, paid subordinates, his position as Reserve Division Commander was about the same as his job in Court Operations.
Nor does the record indicate that Bardzik had significant control over the 600 part-time volunteer reserve members (many of whom were not authorized to wear a badge or carry a firearm). Bardzik explains that as manager of the Reserve Division, his job was to see to it that the reserve members followed already-established departmental policies. The majority acknowledges that Bardzik could not by himself evaluate or promote the reserve members, but was required to send evaluations and recommendations for promotion up the chain of command for approval by the Assistant Sheriff. Maj. op. at 1148-49. Bardzik did not supervise or train the reserve members; instead, they were supervised and trained by the divisions to which they were allocated. The process of allocating reserve members to divisions that needed them was effectively taken out of Bardzik’s hands by the decentralization program; reserve liaisons in each division could request reserve members without notifying Bardzik. Personnel issues were referred to and handled by the personnel division, not by Bardzik. Making every reasonable inference in Bardzik’s favor, it would appear that Bardzik had little control over the reserve component.
The majority tells us that whether Bardzik actually trained or dealt with the personnel issues of his subordinates is immaterial, so long as he “oversaw reserve training and personnel issues.” Maj. op. at 1148-49. The term “oversaw” suggests that Bardzik delegated the training and personnel issues while retaining supervisory authority: There is no evidence in the record that Bardzik retained any supervisory authority over personnel issues once he referred a reserve member to the personnel division. As for training, to the extent that Bardzik was authorized to recommend a course of training for a reserve member, he was still required to obtain the Assistant Sheriffs approval.
The majority cites Thomas v. Carpenter, 881 F.2d 828 (9th Cir.1989) to support the argument that Bardzik was a policymaker. But Thomas favors Bardzik’s position. In Thomas, we held that, on the pleadings, a sheriffs lieutenant was not a policymaker despite being involved in training and evaluating the department’s high risk entry team. Id. at 829, 832. We reasoned that training and evaluation of subordinates “has no significant relationship to one’s political loyalty.” Id. at 832. So even if Bardzik was involved in the training and supervision of the reserve members, under Thomas that involvement would not weigh in favor of finding that Bardzik was a policymaker. The facts of Bardzik’s case *1154are even stronger than those of the lieutenant in Thomas: the record shows that Bardzik was only tangentially involved in training and supervision. Thus, reading the record in the manner most favorable to Bardzik, one must conclude that he did not have any significant control of others, and therefore the control factor does not weigh in Sheriff Carona’s favor.
2. Influence on Programs
The majority argues that Bardzik had influence on the department’s programs, Maj. op. at 1146-47, and that this is the “ ‘most critical factor’ ” in the policymaker analysis. Maj. op. at 1146 (quoting Walker v. City of Lakewood, 272 F.3d 1114, 1133 (9th Cir.2001)). The majority misreads Walker: influence on programs was the “most critical factor” in Walker because the city of Lakewood “had delegated total control over its fair housing program” to the plaintiff. Walker, 272 F.3d at 1133. There is no indication that influence on programs is per se the “most critical factor” in every case. Certainly the Sheriffs Department had not delegated total control over the Reserve Division to Bardzik, given that Bardzik had to seek approval from his superiors for virtually every action he took.
Moreover, many of the programs mentioned by the majority as evidence of Bardzik’s influence do not support the majority’s position. The majority refers to an audit of the Reserve Division that Bardzik helped conduct. But the majority itself points out that Captain Twellman headed that project, not Bardzik. Maj. op. at 1146-47. Further, the goal of the audit was to bring the reserve members into compliance with already-established standards set by a state commission, not to set new policy. Bardzik apparently initiated some programs to train reserve members in swift water rescue and computers, but as discussed earlier, under Thomas, the training of subordinates is not determinative in policymaker analysis. 881 F.2d at 832. Bardzik also initiated the decentralization program, but this served to reduce his influence, relinquishing much of his supervisory authority to the reserve liaisons and other officials in each of the department’s divisions. On balance, the record clearly indicates that Bardzik did not have any significant influence on the department’s programs.
3. Additional Factors
The majority argues that Bardzik had “particular technical competence to serve as Reserve Division Commander.” Maj. op. at 1147. The only evidence of this factor is Sheriff Carona’s statement that Bardzik had “skills,” and Bardzik’s assertion that the Reserve Division was running better than ever during his tenure as commander. These vague assertions are insufficient to draw an inference that Bardzik had “particular technical competence.”
The majority further argues that Bardzik had “vague and broad responsibilities.” Maj. op. at 1146. Whatever generalities Sheriff Carona may have said to Bardzik about his duties as Reserve Division Commander, the record indicates that Bardzik’s main duties involved monitoring reserve members’ compliance with departmental policies, and ensuring that reserve resources were allocated where needed (this latter duty was diminished by the decentralization program, which allowed reserve members to be allocated without Bardzik’s involvement). Contrary to the majority’s assertion, these are not “vague and broad responsibilities.”
The majority also points out that Bardzik had “frequent and meaningful contact” with the Sheriff, an elected official. Maj. op. at 1147-48. But the record makes clear that this contact happened when *1155Bardzik was first appointed as Reserve Division Commander, and ended when a new Assistant Sheriff was assigned to the division in November 2004. The record is silent as to the extent of Bardzik’s contact with Sheriff Carona once the new Assistant Sheriff came on board. Because every inference must be made in Bardzik’s favor, we must assume that after November 2004 he no longer had “frequent and meaningful contact” with the Sheriff. Thus, in October 2005, when Sheriff Carona decided to retaliate against Bardzik by transferring him to Court Operations, there was no “frequent and meaningful contact” between the two, and that factor does not weigh in favor of Sheriff Carona.
Finally, the majority argues that Bardzik was “in a position to thwart the Sheriffs agenda,” an additional factor not mentioned in Fazio. Maj. op. at 1148. The record, in fact, indicates that Bardzik was quite powerless to thwart the Sheriff. As the majority acknowledges, Bardzik “tried to take badges away from the Sheriffs political allies,” but the Sheriff stopped him from doing so. Maj. op. at 1148-49. Similarly, the Sheriff stopped Bardzik from sending memoranda to the reserve members regarding campaign solicitations. Maj. op. at 1149. With limited ability to communicate with the reserves supposedly under his command, and with no public presence or ability to speak on behalf of the department (see Maj. op. at 1146), Bardzik was hardly in a position to do much political damage to the Sheriff. Perhaps Bardzik criticized the Sheriff openly, but so could any sheriffs deputy, whether policymaker or not. This factor does not support a finding that Bardzik was a policymaker.
Conclusion
Because the record as currently developed does not support the conclusion that Bardzik was a policymaker while Reserve Division Commander, I would hold that our decision in DiRuzza v. County of Tehama, 206 F.3d 1304 (9th Cir.2000), compels us to affirm the district court. In DiRuzza, a sheriffs deputy brought suit after being fired for supporting the Sheriffs opponent in an election. Id. at 1306. As in Bardzik’s case, the qualified immunity question came down to whether DiRuzza was a policymaker or not. Id. at 1313-14. We acknowledged that the record was unclear as to “whether DiRuzza’s job actually entailed policymaking.” Id. at 1314. But at the summary judgment stage, drawing all inferences in DiRuzza’s favor, we concluded that she was not a policymaker, and her superiors would have been well aware of that fact. Id. Thus, DiRuzza’s superiors were not entitled to qualified immunity. Id.
Similarly, I would find that, construing the record in Bardzik’s favor, Bardzik was not a policymaker while serving as Reserve Division Commander. Thus, under DiRuzza, I would hold that Sheriff Carona is not entitled to qualified immunity. I would affirm the district court across the board.