The government has decided (thankfully) to kick the national minimum wage into touch. The hope must be that its panel advises that any national minimum wage high enough to make a meaningful difference to the working poor is a very bad idea. It cannot offer much poverty relief to those in jobs without destroying the opportunity to find work for many more, particularly the inexperienced, the young and those outside cities.The problem is that many of those who find work (mostly in the cities) at the lower end of the wage scales remain poor. The working poor are defined as those earning less than R4,000 a month.The problem is that most of those with jobs in South Africa earn much less than this while many potential workers are unemployed and earn no income.According to a recent study of the South African labour market by Arden Finn for the University of the Witwatersrand, 48% of all wage incomes representing five million workers fell below R4,000 a month in 2015 and 40% earned less than R3,000 a month (2.7million workers out of the total employed of about 13million).story_article_left1The proportion of those earning below R4,000 is much higher in rural areas, higher in agriculture (nearly 90%) and domestic services (95%). In mining, 22% earned less than R4,000 in 2015, while in the comparatively well-paid, well-skilled manufacturing sector about 48% earned less than R4,000 a month.How many would lose jobs and how many would keep them to receive the promised benefits of higher minimum wages would have to be estimated by the panel. They would have to allow for all independent forces at work other than wages that could influence numbers employed.For those who believe that South Africa can repeal the laws of supply and demand for labour, that wages have little to do with what workers are expected to add to business revenues, and so that higher minimums can happen without unhelpful employment effects, there is a question to answer.If a higher national minimum wage can make such a helpful difference to poverty without serious consequences for the unemployed and their poverty, why not set it ever higher?If a national minimum wage of R4,000 a month is not enough to escape poverty, why not double or treble it? They must surely agree that job losses would increase as the distance between current wages and intended minimums widens.Agree that the only way to avoid extra unemployment would be to set minimums close to actual minimum wages, a symbolic gesture.The panel could turn to the relationship between employment, employment benefits and output (measured as value added or contribution to GDP) in the formal sector for evidence that improved employment benefits, for those who keep their jobs, leads to less employment for the rest. GDP has grown while employment has stagnated and unemployment numbers have risen as the potential workforce has grown.story_article_right2Yet real wage bills have grown more or less in line with real GDP. The percentage decline in numbers employed has been less than the percentage increase in employment benefits paid out. And wages and profits have maintained their share of value added.Firms have adapted well to more expensive labour, but the unemployed have not been able to. Their interests should be paramount.A national minimum wage well above market-related rewards will reinforce such trends. It will not be fair to the nonworking poor or promote economic growth - the only known way to truly relieve poverty and raise wages over time.It is the duty of economists to practise tough love - to recognise the trade-offs should a national minimum wage be introduced. We hope the panel will resist politically tempting actions with predictably disastrous effects.If we could eliminate poverty with a wave of the national minimum wage wand, we would have done so long ago.Kantor is chief economist and strategist at Investec Wealth and Investment. The views expressed are the author's and may not necessarily represent those of Investec..