COMPANY SAYS EMPLOYER'S DECISION TO USE ACTUARIAL DATA SHOULD BE GIVEN GREAT WEIGHT.

We understand that you are considering changes to the nondiscrimination rules that were issued last year with respect to compliance under Code Sections 401(a)(4) and 410(b). We strongly support the approach of the Treasury and the Internal Revenue Service in seeking ways to simplify and lighten the burden imposed by these regulations to companies such as Colgate. We believe that such simplification and clarification will make it possible for employers to comply more easily without any adverse effect on plan participants. Colgate has always supported the concept of protection of participants' rights, but the administrative burden imposed by the provisions of the regulations as they stand has proved grueling for us.

We have recently learned of the proposed revenue procedure that the government has issued in connection with Announcement 92-81. We strongly applaud this approach to the proposed substantiation guidelines for nondiscrimination testing. Particularly, we feel that the running of the test on a less than annual basis, and the use of the snapshot data on one representative day of the plan year, will be extremely helpful to Colgate. We urge you to continue simplifying the manner in which responsible employers can comply with these tests.

One specific concern that we have with respect to the proposed revenue procedure is the burden placed on employers who seek to use actuarial valuation data to run the various tests. The procedure, when issued in final form, should make clear that an employer's decision to use actuarial data will be given great weight unless fraud or abuse is involved. It would not be proper for an agent on audit to claim that the employer's decision to use actuarial data rather than precise data was inappropriate, and to make the employer rerun and test using more precise data years after it could be retrieved.

We look forward to other measures from the Treasury and IRS to simplify the matters we are discussing.

Sincerely,