SELYA, Circuit Judge. The defendants, state officials, adamant in their refusal to recognize that discriminatory animus can be proven circumstantially as well as by direct evidence, appeal from an order suspending the operation of termination notices addressed to the plaintiffs, seven seasonal state employees. We affirm.
The circumstances of the case can be succinctly summarized.
1. In November 1990, Rhode Island's incumbent governor, a Republican, was defeated at the polls. His successor, Bruce Sundlun, is a Democrat. Governor Sundlun took office on January 1, 1991.
2. Because parimutuel wagering is permitted at the privately-owned jai alai fronton in Newport, Rhode Island, pursuant to a state license, state employees are assigned to work there. The fronton is open on a seasonal basis. The 1991 jai alai season began on May 2, 1991.
3. Toward the end of April, 1991, the state hired ten persons, including the seven plaintiffs, to fill various non-policymaking positions (e.g., parimutuel monitors, auditors, etc.) based at the fronton.1 All seven plaintiffs had worked there without incident during the previous season. All were Republicans.
4. On April 26, the governor's chief of staff, R. David Cruise (himself a Democratic state senator), demanded that the hiring agency (the Division of Racing and Athletics of the state's Department of Business Regulation) furnish him a roster of positions at the fronton. The list was delivered the same day. It enumerated thirteen positions and contained an explicit statement that: "All but three (3) of the positions listed above are currently filled." The three vacant positions were specified.
5. Senator Cruise, who testified that he read the list but not the explanatory statement, immediately consulted a Democratic general officer and a prominent Democratic state representative concerning potential appointees. On May 1, 1991, the chief of staff transmitted thirteen names to the hiring agency, denominating the job that each person was to perform at the fronton. The plaintiffs' positions were encompassed. Twelve of the newcomers were Democrats. The thirteenth, registered as an "unaffiliated" voter, was Cruise's close personal friend.
6. The plaintiffs were cashiered the next day, just as the jai alai season opened.
On these, and other, facts, the district court concluded that, apart from political spoils, it would be "difficult if not impossible to find any other rational explanation [for the firings]." Applying the familiar four-part test for ascertaining whether a preliminary injunction should be granted, see, e.g., Narragansett Indian Tribe v. Guilbert, 934 F.2d 4, 5 (1st Cir. 1991), the court determined that interim relief was warranted.2 It thereupon issued a restraining order which had the effect, pending a trial on the merits, of reinstating the sacked plaintiffs to the jobs that had been snatched away from them so unceremoniously. This appeal followed.
We linger little over the appellants' importunings. The parties agree that partisan political affiliation was not a legitimate requirement for the due performance of any of the affected jobs; and that, therefore, under principles elucidated by the Supreme Court, the plaintiffs were protected against patronage dismissals by a clearly established First Amendment right. See Rutan v. Republican Party of Ill., 497 U.S. 62, 110 S. Ct. 2729, 2737, 111 L. Ed. 2d 52 (1990); Branti v. Finkel, 445 U.S. 507, 517-18, 63 L. Ed. 2d 574, 100 S. Ct. 1287 (1980); Elrod v. Burns, 427 U.S. 347, 367-68, 49 L. Ed. 2d 547, 96 S. Ct. 2673 (1976). The only question, then, is whether the district court erred in concluding that the plaintiffs were likely to succeed on the claim that their discharges were politically motivated.3
Refined to bare essence, the appellants' argument seems to be that political favoritism must be proved by direct evidence. We disagree. Victims of heavy-handed uses of the spoils system are not limited to redress in only those (relatively rare) instances in which a "smoking gun" can be produced. To the exact contrary, we have held, time and again, that circumstantial evidence alone can support a finding of political discrimination. See, e.g., Estrada-Izquierdo v. Aponte-Roque, 850 F.2d 10, 14-15 (1st Cir. 1988); Kercado-Melendez v. Aponte-Roque, 829 F.2d 255, 264 (1st Cir. 1987), cert. denied, 486 U.S. 1044, 108 S. Ct. 2037, 100 L. Ed. 2d 621 (1988); Rosaly v. Ignacio, 593 F.2d 145, 149 (1st Cir. 1979). Circumstantial evidence is often particularly helpful when, as here, a case turns on a protean issue such as an actor's motive or intent. As the Court wrote in a closely analogous context almost half a century ago:
While objective facts may be proved directly, the state of a man's mind must be inferred from the things he says or does. . . . Courts and juries every day pass upon knowledge, belief and intent -- the state of men's minds -- having before them no more than evidence of their words and conduct, from which, in ordinary human experience, mental condition may be inferred.
American Communications Ass'n v. Douds, 339 U.S. 382, 411, 94 L. Ed. 925, 70 S. Ct. 674 (1950). The test, then, as applied to the plaintiffs' dismissals, reduces to a question of persuasiveness: does the circumstantial evidence, taken as a whole, give rise to a plausible inference of discriminatory animus which, ultimately, possesses enough convictive force to persuade a rational factfinder that the defendants' conduct was politically motivated?
In the case at bar, the circumstantial evidence is damning and, consequently, the inference of discrimination is virtually inescapable. To start with, all the plaintiffs were registered Republicans. By and large, they were replaced by Democrats. As the defendants asseverate, these raw numbers alone are likely insufficient to ground a finding of patronage dismissal. See Estrada-Izquierdo, 850 F.2d at 15. But, there was much more.
The timing itself was suggestive: the transmogrification occurred on the heels of a major shift in political power. The evidence showed beyond peradventure that the plaintiffs were qualified for their jobs; indeed, they had performed the same work satisfactorily during the preceding season. They were discharged within a week after they were rehired. To this date, the defendants have not articulated a legitimate reason for the mass firings. Moreover, the dismissals occurred after the governor's chief of staff had inquired specifically about the jobs and had conferred privately with two party leaders.
To be sure, Cruise testified that he had no partisan motives. But, what an actor says is not conclusive on a state-of-mind issue. Notwithstanding a person's disclaimers, a contrary state of mind may be inferred from what he does and from a factual mosaic tending to show that he really meant to accomplish that which he professes not to have intended. On this pregnant record, the district judge acted well within his province in refusing to credit Cruise's claim of wounded innocence. Trial judges, after all, are not required to ignore that which is perfectly obvious or to take a witness' word for his state of mind, no matter how strongly the circumstances point in a different direction. And we, in turn, ought not to disturb supportable findings, based on witness credibility, made by a trial judge who has seen and heard the witnesses at first hand. See Anderson v. City of Bessemer City, 470 U.S. 564, 575, 84 L. Ed. 2d 518, 105 S. Ct. 1504 (1985); Deguio v. United States, 920 F.2d 103, 106 (1st Cir. 1990).
The defendants also argue that, because the evidence fails to show that Senator Cruise knew the plaintiffs or was explicitly aware of the plaintiffs' efforts on behalf of Republican causes, the plaintiffs are unlikely to succeed at trial. This argument is sheer sophistry. Given the nature of the positions, the defendants' knowledge that the plaintiffs had originally been hired by the previous (Republican) administration, the timing of the moves, the identities of those consulted, the lack of any legitimate reason for ousting the incumbents, and the partisan connections of the replacement workers, it seems disingenuous to suggest that Cruise acted without regard to the politics of the situation. We do not think that liability in a political discrimination case involving non-policymaking positions necessarily depends on a finding that the defendants knew to a certainty that the ousted jobholders were members of the opposition party. Cf., e.g., Matlock v. Barnes, 932 F.2d 658 (7th Cir.) (demotion of non-policymaking municipal employee held to be impermissible act of political reprisal; demoting official and demoted employee belonged to same political party, but employee had supported losing candidate in party primary), cert. denied, 116 L. Ed. 2d 247, 112 S. Ct. 304 (1991).
We need go no further. The district court, on the basis of an impressive array of circumstantial proof, determined that the plaintiffs were likely to succeed in this political discrimination suit. At the same time, the court determined that the other factors conducive to the issuance of a preliminary injunction were present. Those findings pass muster on three salient levels: as a matter of law; as a matter of evidentiary sufficiency; and as a matter of common sense. The restraining order that ensued was an appropriate exercise of the court's discretion.
1. Most of these individuals were hired on April 24, 1991.
2. The four factors to be considered are (1) the likelihood of success on the merits, (2) the potential for irreparable injury, (3) the interplay among the relevant equities, and (4) the effect on the public interest. See Guilbert, 934 F.2d at 5. The district court's calibration of these factors is entitled to considerable deference.
3. While the appellants also question whether the irreparable harm component of the preliminary injunction standard was satisfied, they do so only on the basis that the plaintiffs "have not sufficiently demonstrated a probability of success on the merits." Appellants' Brief at 11. As to the standard's third and fourth components, the appellants do not contend that the district court erred in its assessments.