In good times, civil servants benefit from the pay-trend survey that forms the basis for their annual adjustments - a safeguard for pay and conditions that are envied by the rest of the community. In bad times, it is reasonable, therefore, to expect them to accept that when private-sector pay is going down - as it has been - their annual adjustment should reflect that, too. As we report today, however, staff unionists are challenging the latest survey, which indicates that civil service pay should be cut by 0.96 per cent for the lower salary band, 1.98 per cent for the middle band and 5.38 per cent for the upper band. They argue that the inclusion for the first time of a private company with 15,000 employees - only about 8 per cent of the 185,321 in the survey - might have skewed the result. Therefore, there are doubts all civil service unions will endorse it. The argument is flawed. It amounts to accepting the umpire's decision when it goes their way and disputing it when it does not. Everyone should play by the rules. The government consults the unions on the methodology and procedures of the pay survey - input that private-sector workers do not have when their pay is reviewed. The survey is an effective mechanism for maintaining fair relativity only if pay can go down as well as up. Criticism of the inclusion for the first time of the large employer concerned is not convincing. There is no evidence that as a result, the survey does not reflect the true situation. It could also be argued that previous surveys were not representative because this company was not included. If private-sector pay had been rising and the same company had been setting the pace, it is not hard to imagine the outcry if the government sought to discount its influence on the survey to cap a pay adjustment. That said, the government is not bound to follow the survey findings, since it also takes into account its financial situation, the state of the economy and civil service morale. It has dug deep into its reserves to cushion the community against the economic downturn. Now it has to strike a delicate political balance between maintaining morale - at additional cost - and pay cuts that could depress rates in the private sector. A pay cut would need the approval of lawmakers, who would face the same dilemma. Anticipating the result of the survey, the largest civil service union called for a pay freeze well before it came out. That would be a politically appealing way out for the government. But the case for a freeze is convincing only for the lower-paid tier that can least afford a pay cut, since the adjustment indicated by the survey is very small. It is harder to argue that better-paid civil servants should not share a little of the pain with the rest of the community. Come what may, they still have the ultimate protection of job security. If the annual pay-trend survey is to have any public credibility in good times or bad, the civil service unions should accept the outcome of a mechanism they have agreed to.