5 ms·

It is risky to use econometrics as even qualitative data to support a hypothesis about the trajectory of history over millennia because the data doesn't exist.

It is risky to use econometrics as even qualitative data to support a hypothesis about the trajectory of history over millennia because the data doesn't exist. If you're not extremely careful in analyzing the provenance of your source statistics, it's very easy to fall into a circular argument of "assuming my hypothesis, my hypothesis is true." Here, just staring at the GDP per capita data, for example, the source paper pretty bluntly says that it can't back-continue the statistics past 1820, so it uses a completely different methodology to try to offer a number. So the author here commenting that the data is quite different pre- and post-Industrial Revolution shouldn't be surprising considering the data is literally based on completely different metrics! Hell, De Long literally shaves a factor of 4 off pre-Industrial Revolution GDP estimates because he feels like it.

mr_toad 2y ago
I don’t think long-term GDP measures are meaningful. What point is there in comparing an economy that produces horseshoes and wooden ships with one that has trains and steamships? A pre-industrial economy can’t produce X% of an industrial economy, because it can’t produce large chunks of that economy at all. At best you can compare the parts of the economy that remain relatively unchanged - like food, or textiles, and use those as rough and incomplete proxies.
- eru 2y ago
  When both economies are existing at the same time, and even trading with each other, you can relatively easily compare them. Eg you can compare Britain that had already (partially) industrialised with her continental peers who hadn't.
cryptonector 2y ago
Life expectancy absolutely is a good metric for comparing the Industrial Revolution to what came before. GDP per-capita less so to some degree, and even today it's difficult to compare 1st world GDP per-capita to places where it's three orders of magnitude slower because GDP measures generally fail to properly capture the production of subsistence farming and other small-scale (but valuable to the people in question) activities . And yet! it is obvious that GDP per-capita has been flatish throughout history by comparison to post Industrial Revolution times precisely because the vast majority of people do not live off subsistence farming, make their own homes, clothing and footwear, etc.
- mr_toad 2y ago
  > And yet! it is obvious that GDP per-capita has been flatish throughout history by comparison to post Industrial Revolution times The early part of exponential growth always looks flat compared to the last part, it’s the proverbial hockey stick. There are actually continuous developments in agriculture through the Iron Age. off the top of my head there are windmills, more and better iron tools, heavy iron plows, improved crop rotation, more advanced irrigation etc. The Arab agricultural revolution changed a lot of farming technology, and facilitated exchanges between east and west. Then there was the post Columbian trade with the new world. Imagine a world without potatoes!
  - HPsquared 2y ago
    Not forgetting the social and legal structures of these activities. The joint-stock company for instance. Also just the terminology: water-mills, windmills and sailing ships are definitely industrial equipment, even if they're not steam-powered.
- jcranmer 2y ago
  > it is obvious that GDP per-capita has been flatish throughout history But is it? That's ostensibly the point of this article--trying to actually use data to establish whether or not this "obvious" principle is true [1]. But if you turn to the actual data, the author of the data is creating the historical data using one of three assumptions: either real GDP per-capita is correlated with population growth, which comes both with and without an arbitrary pre-industrial haircut, or real GDP per-capita is literally flat pre-industrial. It really isn't clear to me if there really is a step change in per-capita GDP growth around the First (and/or Second) Industrial Revolution, or if it's a one-off jump, or if it's a continuous, consistent per-capita growth rate with no real impact. All of those would look pretty similar in a linear-scale graph at first glance. What we do know from other economic proxies from archaeology is that pre-industrial economies weren't exactly static: you can see a clear trend that highlights the rise and fall of Rome (see, e.g., the data in https://acoup.blog/2022/02/11/collections-rome-decline-and-fall-part-iii-things/ https://acoup.blog/2022/02/11/collections-rome-decline-and-f...), so you better expect that the econometric statistics you use can replicate that trend. [1] Let's assume the purposes of discussion that GDP here is something meaningful for economic prosperity in pre-industrial times, something that can capture the productivity enhancements that did occur at that time.

mr_toad 2y ago

I don’t think long-term GDP measures are meaningful. What point is there in comparing an economy that produces horseshoes and wooden ships with one that has trains and steamships? A pre-industrial economy can’t produce X% of an industrial economy, because it can’t produce large chunks of that economy at all. At best you can compare the parts of the economy that remain relatively unchanged - like food, or textiles, and use those as rough and incomplete proxies.

eru 2y ago
When both economies are existing at the same time, and even trading with each other, you can relatively easily compare them. Eg you can compare Britain that had already (partially) industrialised with her continental peers who hadn't.

cryptonector 2y ago

Life expectancy absolutely is a good metric for comparing the Industrial Revolution to what came before. GDP per-capita less so to some degree, and even today it's difficult to compare 1st world GDP per-capita to places where it's three orders of magnitude slower because GDP measures generally fail to properly capture the production of subsistence farming and other small-scale (but valuable to the people in question) activities . And yet! it is obvious that GDP per-capita has been flatish throughout history by comparison to post Industrial Revolution times precisely because the vast majority of people do not live off subsistence farming, make their own homes, clothing and footwear, etc.

mr_toad 2y ago
> And yet! it is obvious that GDP per-capita has been flatish throughout history by comparison to post Industrial Revolution times The early part of exponential growth always looks flat compared to the last part, it’s the proverbial hockey stick. There are actually continuous developments in agriculture through the Iron Age. off the top of my head there are windmills, more and better iron tools, heavy iron plows, improved crop rotation, more advanced irrigation etc. The Arab agricultural revolution changed a lot of farming technology, and facilitated exchanges between east and west. Then there was the post Columbian trade with the new world. Imagine a world without potatoes!
- HPsquared 2y ago
  Not forgetting the social and legal structures of these activities. The joint-stock company for instance. Also just the terminology: water-mills, windmills and sailing ships are definitely industrial equipment, even if they're not steam-powered.
jcranmer 2y ago
> it is obvious that GDP per-capita has been flatish throughout history But is it? That's ostensibly the point of this article--trying to actually use data to establish whether or not this "obvious" principle is true [1]. But if you turn to the actual data, the author of the data is creating the historical data using one of three assumptions: either real GDP per-capita is correlated with population growth, which comes both with and without an arbitrary pre-industrial haircut, or real GDP per-capita is literally flat pre-industrial. It really isn't clear to me if there really is a step change in per-capita GDP growth around the First (and/or Second) Industrial Revolution, or if it's a one-off jump, or if it's a continuous, consistent per-capita growth rate with no real impact. All of those would look pretty similar in a linear-scale graph at first glance. What we do know from other economic proxies from archaeology is that pre-industrial economies weren't exactly static: you can see a clear trend that highlights the rise and fall of Rome (see, e.g., the data in https://acoup.blog/2022/02/11/collections-rome-decline-and-fall-part-iii-things/ https://acoup.blog/2022/02/11/collections-rome-decline-and-f...), so you better expect that the econometric statistics you use can replicate that trend. [1] Let's assume the purposes of discussion that GDP here is something meaningful for economic prosperity in pre-industrial times, something that can capture the productivity enhancements that did occur at that time.