Thinking about macro-economics hurts my head. I can’t explain why without getting pretty technical, so feel free to skip this post if you’re not a fan of calculus.
The economy, measured by GDP, is often thought of as a being a stock, and is described as growing and shrinking. When GDP grows, the economy is booming, which is good. When GDP decreases, the economy is in recession or depression, which is bad. When GDP growth is smaller over a period of time, that’s stagnation, which is also bad.
This all makes sense if you think of GDP as a stock, but it’s not. It’s a flow. (Click here for groovy graphics about stocks and flows) Human wealth, which is actually the most important thing, is a stock. GDP affects wealth; it’s an input to wealth’s first derivative. Change in GDP affects wealth; it’s an input to wealth’s second derivative. But the conversation about the strength of the economy should start with wealth, not GDP.
In order for wealth to increase, GDP growth does not need to be high. GDP growth doesn’t even need to be positive. So long as GDP exceeds depreciation, wealth increases over time. That is, so long as we produce more value in a year than the annual decrease in value of our possessions, we get wealthier. Ideally, wealth not only increases, but grows as a quadratic function, which occurs when GDP growth is positive. And even more ideally, wealth grows as a juicy quadratic function, rather than a less juicy quadratic function, which occurs when GDP growth is high. But that’s icing on the cake.
Have I made your head hurt yet? No? Read on.
When Tyler Cowen links favorably to this post by David Beckworth, which uses this chart, I ask myself, and David Beckworth and Tyler Cowen, why are you using a logarithmic scale? If Total Factor Productivity is an input to the first derivative of wealth, and it’s grown steadily, then wealth has grown as a quadratic equation. If you display the growth of TFP on a logarithmic scale, it will appear to slow down, because that’s what logarithmic scales do to linear growth rates. That doesn’t mean it’s an actual problem.