“…Between 1994 and 2008 there was weak employment growth, mostly in low-paying semi-formal 'jobs' and informal work (Casale, Muller and Posel, 2004;Bhorat and Oosthuizen, 2006) -at least until 2008-09, when about one million jobs were lost. PALMS has improved the consistency of wage data, showing, inter alia, that the purported 'collapse' in earnings the late 1990s (as reported by Casale, 2004;Banerjee et al, 2006;Leibbrandt, Levinsohn and McCrary, 2010; and, based on these, Altman, 2006;Marais, 2011) was 'almost definitely an artefact of changing survey practice', especially sampling and the enumeration of informal sector activity (Wittenberg, 2014: 60).…”