The flaw with traditional economic indicators and what to use instead.

Open up any macro/economic reports and you will likely find commentaries based on broad indicators such as GDP trends, unemployment rates, stock market levels, inflation, money supply, balance of trades blah blah. These are okay except that they come with three issues:

1. They are not intuitive:

For instance, take the media's favorite unemployment rates. An unemployment rate of 5% does not mean 5% of the total population is unemployed. No ma’am (sir). Labor force participation measures the percentage of population who are presently employed or seeking employment and unemployment rate in turn measures the percentage within that labor force that are actively seeking but unable to find work. India's average labor participation rate was ~54% between 2005-18. A 5% unemployment rate means 5% of the 54% were unemployed.

Yet another is metric is the stock market level; which contrary to popular belief is not a very robust financial barometer of the economy. High (or low) market levels occur not because the economy is faring well (or poorly) but primarily due to speculative trading. As was recently evident during the April – June period, when the Sensex was rising even though the country was in lock-down.

2. Suffer from estimation errors:

This is common to almost all indicators. However, broad indicators such as unemployment rates, inflation, GDP are more prone to higher sampling errors. For example, inflation is measured based on a certain basket of goods. Also, inflation is differs according to consumer or wholesale level and urban and rural level. Other indicators such as unemployment rates are measured via a sample surveys and the inputs to GDP computation are estimates in the first place etc.

3. Lack comparability:

Broad indicators do not have good comparability. For example, it’s not fair to compare GDP growth rates of China with India. It’s a much bigger deal to grow $15 trillion economy (China) at 6.7% than a $2 trillion economy (India) at 7.2%. Period.

Also, let's visit unemployment rates again. Unemployment rate indicates percentage of unemployed and employed persons but it does not point out the quality or pay-ability of jobs. On paper, we might have better scores than some countries, but the quality and pay of our employment may suck.

Also, broad indicators are lagging that is they measure production and performance only after they have occurred. They confirm long-term trends but do not necessarily predict them.

So what to use instead?

I find it useful complementing traditional indicators with a mix of proxy indicators. Proxy indicators include metrics such as the sales of two-wheelers, sales of commercial vehicles, sales of agriculture equipment (like tractors, pumps etc.), airline passenger traffic, railway freight traffic, consumption per person (electricity, petroleum, steel, cement), export-import deficit, credit growth, no. of new business registrations, no. of new construction permits etc. which are positively correlated with economic well-being.

These are instantaneously more intuitive as they exactly mean what they mean and do not have complex interpretations such as GDP, money supply etc. Also, they are comparable among peers and suffer less from estimation errors as they lend themselves to simple aggregations and validation from multiple sources; i.e is any one institution can't fudge up the numbers. The total number of two-wheelers can be aggregated if you add up unit sales of all two-wheeler companies selling in the country.

And that's that.