The marginal value of government

Tyler Cowen provides some food for thought in The Great Stagnation:

Let's say government spends $1 million fixing a road: How much does that contribute to measured GDP? $1 million. No consumer "buys" the road, but the expenditure counts nonetheless toward the output of goods and services. In other words, in measured GDP, we are valuing the expenditure at cost.

...Sometimes government outputs are worth a lot more than what we spend them on, sometimes they are worth a lot less...A dollar spent on very basic police and courts and army protection is worth more than a dollar spent refurnishing a warehouse in Minneapolis under the guise of urban renewal. A dollar spent on welfare for the poorest is more valuable than a dollar spent extending the program to better-off but still poor cases. And so on. Yet when it comes to measuring GDP, we are valuing every one of these different expenditures at $1. In our measurements, we are assuming that the quality, importance, and efficacy of government stays constant as the size of government grows.

...Thus, usually, when we spend another dollar through government, it is worth a bit less -- on average -- than the last dollar we spent on government. Government, at the margin, is becoming less productive. Yet, when measuring GDP, we treat each dollar of government spending as if it is equal in value to the previous dollars that were spent. We're valuing dollars spent on highways extensions as if they were worth as much as the dollars we spent on building the core roads that link major cities.

After laying this groundwork, Cowen gets to his main points:

To better measure how well we are doing as a nation, remember this about productivity: The larger the role of government in the economy, the more the published figures for GDP growth are overstating improvements in our living standard.

...There is a corollary, namely: The larger the percentage of government consumption in the economy, the harder it is to tell exactly how well we are doing in real economic growth and living standards.

What a great insight. Government suffers from diminishing marginal utility that is not reflected in GDP stats, and therefore the bigger component government spending is of GDP, the less accurately it reflects the state of well-being. In other words, if country A and country B have identical per capita GDP levels, but government spending comprises 20% of GDP in country A and 40% in country B, country A is clearly better off. The first thing this brings to mind is the European welfare states.

I hope to have more to say about this book in the days ahead.