Richard Murphy says:
June 5 2017 at 11:55 am
Not at all

If you think data can answer a social science question in isolation you don’t understand the question

Or the people it impacts

Which is why most of the real economic greats never went near a correlation

They did not need to


The truth is,” says Mr. Keynes, “that sensible investigators only employ the correlation coefficient to test or confirm conclusions at which they have arrived on other grounds.”

19 thoughts on “Sigh”

  1. The real economic great wrote before correlation coefficients had been invented.


    …they’re kind of saying the same thing…

    Re-read both statements. They are not saying the same thing… kind of or otherwise.

  2. Social Justice Warrior

    If you think data can answer a social science question in isolation you don’t understand the question

    Murphy’s point that we should be cautious about interpreting correlations is correct. And the data he’s talking about – showing a small negative correlation between minimum wage and unemployment – are a good example.

  3. Has social science ever answered a question?

    The only possible interpretation of any research whatever in the ‘social sciences’ is: some do, some don’t.
    Ernest Rutherford (Baron Rutherford of Nelson) 1871-1937

  4. SJW yet the fact that there is no correlation means that Snippa feels that his conclusion is justified. Despite the data not being strong enough to support, he still thinks that mimimum income levels have no impact on unemployment. Economic theory would suggest that they do have an impact. He resorts to bashing academic economics rather than providing any explanation that makes sense

  5. In fact his conclusion is that higher minimum wages are good for unemployment.. This is certainly not a conclusion that is obvious from the data.

  6. Diogenes said:
    “SJW yet the fact that there is no correlation means that Snippa feels that his conclusion is justified”

    It’s worse than that; the evidence actually says that Murphy is wrong, although apparently not very strongly. But Murphy insists that the evidence is wrong, not his theory.

  7. Funny how everyone seems to have read the correllation in the opposite direction I did. My immediate thought was; a society that increases minimum wage is likely to be a society with falling unemployment. Why would one increase the minimum wage if that’d cause a rise in unemployment? You’d have to be barking.

  8. Social Justice Warrior

    the evidence actually says that Murphy is wrong

    Look again: the evidence suggests that higher minimum wage is correlated to lower unemployment. Murphy is showing intellectual honesty by stressing that the evidence is weak.

    Economic theory would suggest that they do have an impact.
    It’s unclear. Making it more expensive to employ people would tend to reduce employment. However, increasing the income of minimum-wage employees would tend to increase growth and hence increase employment, since poor people spend their income.

    But my guess is that what the graph is actually telling us is that governments are more likely to increase the minimum wage if unemployment is relatively low.

  9. SJW my impression for the UK is that the minimum wage has made it difficult for employers to take on people who are at the limits of their value-added – such as people with physical or mental disabilities. It has probably also had an effect on the number of people on zero hour contracts. My evidence is anecdotal but the local supermarket used to employ mentally handicapped guys to round up the trolleys in the car park. They seem to have reduced staffing levels lately and that might have something to do with wage levels but…

  10. If the minimum wage is dependent upon unemployment levels, shouldn’t the regression have been done the other way round?

  11. The problem with any analysis posed along these lines is that Greece and Spain are at the average level in terms of minimum wage but unemployment levels are off the scale. Therefore this analysis is not going to tell us anything much. Comparative data if those 2 countries cut minimum wage rates might be better but surely a more sophisticated analysis is required. Point data tells us little :we need a time series

  12. According to Murphy the only “Economic Greats” are those who agree with him.

    @ SJW I have pointed out on so many occasions that I have lost count that the result of Blair’s introduction of a National Minimum Wage to help lady immigrants working in sweatshops in the textile industry in Leicester was that *over 90%* of workers in the British textile industry lost their jobs.
    I could also mention the loss of Remploy which, when I was young, was providing a valuable service and some good quality simple furniture – but, being intellectually honest, I recognise that was down to appalling management.

  13. So Much For Subtlety

    Chris Miller – “Has social science ever answered a question?”

    Sure. Just not in a way that anyone can reproduce. It is a reasonable assumption that everything in the social sciences is untrue. Virtually nothing can be reproduced. Indeed there are only two robust findings in social science.

    One is that IQ matters and there are large differences between the races

    The other is that pretty much every stereotype you can think of is true.

    This may explain why everything else is social science is rubbish.

  14. Social Justice Warrior

    John77: repeating stuff doesn’t make it true. UK apparel manufacture was in steep decline before the minimum wage was introduced. Sweatshops in Leicester pay about half minimum wage, but it’s still difficult to compete with workers in Bangladesh being paid a few cents an hour.

    Do you want to make UK coal mining competitive by repealing the 1842 Mines Act?

  15. @ SJW
    Go, look at the data!
    It isn’t true just because I repeat it – I repeat that because it IS true.
    Incidentally have you heard about the existence of transport costs? and transport times? A lot of Leicester’s business was short runs when Next ran out of one unexpectedly popular garment and needed some more quicker than they could be shipped from China (it was usually China, not Bangladesh), but Nottingham Manufacturing was destroyed when M&S stopped buying British because the NMW made the difference price that it would have to charge over Next’s too big. Once NMW was introduced Next chose to over-order and take a reduced risk of running out as the cost of topping-up from UK sources exceeded the loss due to having to sell off more at cut prices in the end-of-season sales.

    “repeating stuff doesn’t make it true.” The steep decline was after 1997 – the decline between 1987 and 1997 wasn’t 90%. I’ve heard the claim about the “steep” decline before and *I’ve checked it out* – it was NOT anywhere like as steep.

    UK coal mining always paid more than surface labour, usually a lot more. I don’t want any UK coal mining but NMW is totally irrelevant to coal mining, as you *should* know..

Leave a Reply

Your email address will not be published. Required fields are marked *