Clear evidence? It looks like Reps DIngell and Maloney use the stats in the same way as drunks use lamp posts - for support rather than enlightenment. Without additional information, these numbers can be quite misleading.
First, they are "snapshots" not longitudinal data, they do not take work history into account. For example, if more young women with little work experience entered the labor market in 2000 than in 1995 - that would skew the 2000 ratios downward without implying any discrimination. Au contraire, they might (but do not have to) suggest greater employment opportunity for women in 2000 than in 1995.
Second, grouping wages by industrial categories lumps together people working in the same industries but in different capacities. Again it may indicate discrimination or mere segmentation into niches, but to tell which you would need to disaggregate these industries (e.g. into three digit SIC-codes). A less misleading way of presenting these data would by occupational categories, or a combination of occupational categories and industries (e.g. RN's in hospitals).
Third, it is unclear whether these figures take into account the number of hours worked. For example, if women work fewer hours than men, their wages will naturally be lower. Again, that may or may not indicate discrimination - but we do not know until we examine the circumstances leading to part time employment.
Fourth, these figures do not account for work experience and qualifications (we need longitudinal data for that). If women work experience is, say, shorter than that of men - that will almost certainly be reflected in wages, because seniority is one of the key factors affecting promotions and wages. That does not necessarily indicate discrimination, especially by employers. It could be societal gender roles and expectations, or simply the career vs. family choice.
wojtek