In a perfectly free and open market economy, the type of employer—government or private—should have little or no impact on the earnings differentials between women and men. However, if there is discrimination against one sex, it is unlikely that the degree of discrimination by government and private employers will be the same. Differences in the degree of discrimination would result in earnings differentials associated with the type of employer. Given the nature of government and private employers, it seems most likely that discrimination by private employers would be greater. Thus, one would expect that, if women are being discriminated against, government employment would have a positive effect on women's earnings as compared with their earnings from private employment. The results of a study by Fuchs support this assumption. Fuchs' results suggest that the earnings of women in an industry composed entirely of government employees would be 14.6 percent greater than the earnings of women in an industry composed exclusively of private employees, other things being equal. In addition, both Fuchs and Sanborn have suggested that the effect of discrimination by consumers on the earnings of self-employed women may be greater than the effect of either government or private employer discrimination on the earnings of women employees. To test this hypothesis, Brown selected a large sample of white male and female workers from the 1970 census and divided them into three categories: private employees, government employees, and self-employed. (Black workers were excluded from the sample to avoid picking up earnings differentials that were the result of racial disparities.) Brown's research design controlled for education, labor-force participation, mobility, motivation, and age in order to eliminate these factors as explanations of the study's results. Brown's results suggest that men and women are not treated the came by employers and consumers. For men, self-employment is the highest earnings category, with private employment next, and government lowest. For women, this order is reversed. One can infer from Brown's results that consumers discriminate against self-employed women. In addition, self-employed women may have more difficulty than men in getting good employees and may encounter discrimination from suppliers and from financial institutions. Brown's results are clearly consistent with Fuchs' argument that discrimination by consumers has a greater impact on the earnings of women than does discrimination by either government or private employers. Also, the fact the women do better working for government than for private employers implies that private employers are discriminating against women. The results do not prove that government does not discriminate against women. They do, however, demonstrate that if government is discriminating against women, its discriminating is not having as much effect on women's earnings as is discrimination in the private sector. The passage mentions all of the following difficulties that self-employed women may encounter EXCEPT ______.