This post may turn out to be unsatisfying, because it will ask some questions to which there are no good answers.  We know that if we want something to improve, we have to measure it.  So we measure the nation’s GDP (Gross Domestic Product) and its growth (or shrinkage in a recession).  And if we divide the GDP by the population, we get GDP per capita.  If it is growing, we say that the standard of living is improving.

GDP is the sum of Consumption spending, Investment spending, and Government spending.  In the relatively short run, measurement of changes in GDP per capita are probably fairly accurate. There are some problems with it, because it measures dollars spent, rather than value received, but one year over another is probably not too far off.  The problem I see is over longer periods, where technological change may deliver far more value than just the number of dollars spent.  Let me give a few examples of what I mean.

If someone spends $40 per month on his cellphone service, and increases it to $50 per month, that will increase GDP by $10 per month.  But if another person, who had no cell phone, cancels a land line that costs $40 per month, and subscribes to a cellphone plan at $40 per month, there is no change in GDP.  However, the increase in value received is enormous.  Free long distance calls (up to some limit), and the mobility of communications all beat the static nature of landlines and extra charges per long distance call.  GDP measurements won’t show this.

If XYZ company had 10 million subscribers at $40 per month each, drops the rate to win new subscribers to $30 per month, and gains 2 million more subscribers here’s what happens.  Before the price cut, they had revenues of $400 million per month, and afterwards, only $360 million per month.  But there was an increase of 20% in the people using cell phones, gaining the value they provide.  GDP went down, value provided went up.

Other key statistics have problems, too.  We like to see a low unemployment rate, below 5%.  But that doesn’t measure the participation in the labor force, which may be dropping at the same time the unemployment rate is dropping, because people give up looking for a job, quit the labor force, take government transfer payments, and are no longer counted as unemployed.

As we know, the government can borrow money long term, and spend it today.  GDP may rise through increased government spending (unless it’s for transfer payments which aren’t counted in GDP).  But we increase today at the expense of the future, when our children or grandchildren have to pay the debt through taxes.  Some will argue that debt can always be increased, we won’t go bankrupt.  That is true only up to a point—remember the problems Greece had with a sudden inability to borrow as lenders lost confidence.

So, what, if anything, is the lesson here?  I think it is that no single statistical measure can really tell us how we are doing.  Several differing measures must be looked at simultaneously, to get a whole picture that is relevant.  Politicians tend to only cite the statistic that bolsters the argument they are making at any given moment.  Policies must be designed that truly make our situation better, not just a single statistic.  As Mark Twain said, “Statistics don’t lie, but liars can use statistics”.