Ensuring consistency between confidential data value types

1. A system comprising:

a non-transitory computer-readable medium having instructions stored thereon, which, when executed by a processor, cause the system to:

obtain an anonymized set of confidential data values of a first confidential data type;

obtain an anonymized set of confidential data values of a second confidential data type;

determine a multiplier following a log-normal distribution for the anonymized set of confidential data values of the first confidential data type;

perform smoothing independently for the anonymized set of confidential data values of the first confidential data type and the multiplier, wherein the smoothing comprises modification of at least one confidential data value of the first confidential data type and the multiplier; and

compute percentiles for the anonymized set of confidential data values of the second confidential data type using the smoothed anonymized set of confidential data values of the first confidential data type and the smoothed multiplier.

2. The system of claim 1 , wherein the computing comprises, for each percentile, multiplying the smoothed anonymized set of confidential data values of the first confidential data type by the smoothed multiplier.

3. The system of claim 1 , wherein the first confidential data type is a subset of the second confidential data type.

4. The system of claim 3 , wherein the second confidential data type is total compensation and the first confidential data type is a component of total compensation.

5. The system of claim 4 , wherein the first confidential data type is base salary.

6. The system of claim 4 , wherein the first confidential data type is tip income.

7. The system of claim 4 , wherein the first confidential data type is stock compensation.

8. A computerized method, executable by a hardware processor, comprising:

obtaining, by the hardware processor, an anonymized set of confidential data values of a first confidential data type;

obtaining, by the hardware processor, an anonymized set of confidential data values of a second confidential data type;

determining, by the hardware processor, a multiplier following a log-normal distribution for the anonymized set of confidential data values of the first confidential data type;

performing, by the hardware processor, smoothing independently for the anonymized set of confidential data values of the first confidential data type and the multiplier, wherein the smoothing comprises modification of at least one confidential data value of the first confidential data type and the multiplier; and

computing, by the hardware processor, percentiles for the anonymized set of confidential data values of the second confidential data type using the smoothed anonymized set of confidential data values of the first confidential data type and the smoothed multiplier.

9. The computerized method of claim 8 , wherein the computing comprises, for each percentile, multiplying the smoothed anonymized set of confidential data values of the first confidential data type by the smoothed multiplier.

10. The computerized method of claim 8 , wherein the first confidential data type is a subset of the second confidential data type.

11. The computerized method of claim 10 , wherein the second confidential data type is total compensation and the first confidential data type is a component of total compensation.

12. The computerized method of claim 11 , wherein the first confidential data type is base salary.

13. The computerized method of claim 11 , wherein the first confidential data type is tip income.

14. The computerized method of claim 11 , wherein the first confidential data type is stock compensation.

15. A non-transitory machine-readable storage medium comprising instructions, which when implemented by one or more machines, cause the one or more machines to perform operations comprising:

obtaining an anonymized set of confidential data values of a first confidential data type;

obtaining an anonymized set of confidential data values of a second confidential data type;

determining a multiplier following a log-normal distribution for the anonymized set of confidential data values of the first confidential data type;

performing smoothing independently for the anonymized set of confidential data values of the first confidential data type and the multiplier, wherein the smoothing comprises modification of at least one confidential data value of the first confidential data type and the multiplier; and

computing percentiles for the anonymized set of confidential data values of the second confidential data type using the smoothed anonymized set of confidential data values of the first confidential data type and the smoothed multiplier.

16. The non-transitory machine-readable storage medium of claim 15 , wherein the computing comprises, for each percentile, multiplying the smoothed anonymized set of confidential data values of the first confidential data type by the smoothed multiplier.

17. The non-transitory machine-readable storage medium of claim 15 , wherein the first confidential data type is a subset of the second confidential data type.

18. The non-transitory machine-readable storage medium of claim 17 , wherein the second confidential data type is total compensation and the first confidential data type is a component of total compensation.

19. The non-transitory machine-readable storage medium of claim 18 , wherein the first confidential data type is base salary.

20. The non-transitory machine-readable storage medium of claim 18 , wherein the first confidential data type is tip income.