单项选择题

    _1_ job applications to eliminate discrimination is not easy. Ten big employers in the public and private sectors—including the civil service, HSBC and Deloitte—have agreed to start _2_ on a “name-blind” basis in Britain; others may also follow suit. In such schemes, those drawing up shortlists of applicants cannot see their _3_ , with the aim of _4_ racial and sexual bias. But do they work
    In France a law passed in 2006 made the anonymising of applicants’ CVs _5_ for firms of over 50 employees. But the government was slow in laying _6_ the conditions for how the law would operate, and only started _7_ it last year. In Sweden and the Netherlands there have been some trials. Discrimination  against  job  applicants  based  on  their  names  is  well _8_,  particularly  among ethnic minorities. An experiment in Germany found that candidates with German-sounding names were 14% more likely to be called _9_ an interview than candidates with Turkish ones. A review of various studies, by the Institute for the Study of Labour (IZA) , a German outfit, found that anonymised job applications _10_ the chances of ethnic-minority candidates being invited to a / an _11_ . A Swedish study found that it led to more _12_  candidates being  hired.
    However, the results from other trials are not _13_ . A second Swedish experiment found that only women, not immigrants, were boosted by anonymous recruitment. According to the IZA, experiments in the Netherlands showed no _14_ in the  likelihood  of  ethnic-minority  candidates being offered a job if their CVs were seen anonymously, suggesting that discrimination had _15_ at the interview stage.
    Ensuring that a candidate is completely _16_ is also tricky. A 2012 French study found that _17_ candidates and those from poor districts were less likely to be called for interview when applications were anonymised. Its authors suggested that recruiters may have used other _18_ , such as knowledge of Arabic, to _19_ race.
    Going name-blind when shortlisting candidates may be a sensible start, but it is likely to be just a small step towards _20_ hiring bias.