> Farms generally have increasing returns to plot size.
The status of the IFSP - or rather, the IFSPs, because of course there are many productivity measures you might care about here - is probably the oldest open question in development economics, rounding it off to "generally true" or "generally false" without further qualification is not reasonable.
> the second phase (which essentially privatized Japanese state-owned lands)
This is not fully accurate. From Kim and Wang:
"The second phase of land reform, passed in 1951, redistributed most public lands—around
20% of all arable land on Taiwan (C. Chen 1961). Most of these lands were confiscated from private Japanese colonists expelled from Taiwan—in particular, the four major Japanese sugar corporations, which the Nationalists consolidated into the single Taiwan Sugar Company"
They continue
"Taiwan’s relatively strong system of land rights, intended to win the support of local landlords
for the colonial regime, made it difficult for sugar companies to amass land through coercion,
as the Dutch did on Java. Instead, the Japanese left the basic structure of smallholder farming
intact, contracting with these small farms for cane but maximizing profits by gaining control of
the milling, refining, and marketing of sugar (Cheng, Fan, and Wu 2022). The sugar companies
then used a system of advances to ensure that farmers planted sugar (Williams 1980)."
which seems to me to be a pretty close match for the standard "tenant farming reduces labor intensity" argument in favor of land reform.
Edit: Never mind, it turns out that you already have an article where you talk about this.
Also, have you ever heard of Silvio Gesell and his ideas? I've been trying to get other Wikipedia editors to help me translate the German article into English, and just spread awareness about his monetary reforms in general across Wikipedia. https://en.wikipedia.org/wiki/Silvio_Gesell
I also have to hard disagree on the 'increasing returns claim.
As Amicus points out, this is one of the oldest and most entrenched debates in development economics. I have never heard anyone (except perhaps Soviet planners!) claiming that, historically, there have been 'increasing returns to scale in agriculture' The near consensus view, as of the late 1990s (see e.g. Handbook chapter by Binswanger, Feder, Deininger, 1995) was that, if anything, there were DISeconomies of scale or constant returns to scale in most crops. This explained the ubiquity and dominance of family-based and tenancy farming throughout the world for millenia. Large estates and plantations were generally viewed as arising out of distortions in policy and/or coercion. But there was a lively debate and much empirical investigation. The main issue was to disentangle was whether the commonly observed inverse relationship was due to due to contracting frictions (e.g. principal agent problems) plus technological constant returns to scale or perhaps issues of mismeasurement. Read Binswanger, Feder, Deininger 1995 for a survey of views up to that data (and World Bank policy).
It's only after the 1950s in highly developed countries such as the United States that the idea of increasing returns to scale starts to be entertained. Over time since (with the rise of large-scale farming in places like Brazil and Argentina) this had to be increasingly entertained. The World Bank's 2008 World Development Report (on Agriculture for Development) tackled these issues in depth, and for the first time acknowledges that technology had changed enough to question the old consensuses.
People cite the papers by Restuccia et al as if that had settled the debate. They bring some new modeling and evidence to the table, but their results have hardly gone unquestioned.
Ironically, the Restuccia et al papers build off Lucas' 1978 'On the size distribution of firms" paper which starts from the assumption that there are DISeconomies of scale in production, or more precisely, that there are constant returns to scale, but one factor such as managerial ability is non-traded due to contracting problems (exactly the assumption of the earlier ag literature!). So how do Adamopolous and Restuccia 2014 reach the conclusion that there are what look like economies of scale in production, and therefore 'too many small farms' in the world, supposedly imposing crippling inefficiencies? By slipping in decidedly weird assumptions about the shape of the production function (farms become less not more capital intensive as they grow in land size in ways that lower their costs). This is justified by supposedly calibrating such a function to the farm size distribution to the US economy in an exercise that lumps large grazing farms in Wyoming with wheat farms in Iowa or New England to get a farm size distribution with very large farms. They then assume that the rest of the world should follow this pattern and when, unsurprisingly, the world doesn't they declare this as evidence of huge misallocations. Gollin and Udry have paper in the 2021 JPE challenging the use of this approach and some of their more micro-studies using data from Malawi, by demonstrating that much of what these papers label as evidence of misallocation, is almost due to mismeasurement.
None of this is to say that it's not true that there very likely are "too many small farms" in the world, or that agriculture could and should become much more productive. There are plenty of reasons to believe that many farms fall below minimum efficient scale. But a U shaped AC curve is not evidence of IRS!
The debate over land reform is old and fraught. There have been many long and thoughtful takes on the subject by Alain de Janvry (lead author on that WDR report) and many others.
> Farms generally have increasing returns to plot size.
The status of the IFSP - or rather, the IFSPs, because of course there are many productivity measures you might care about here - is probably the oldest open question in development economics, rounding it off to "generally true" or "generally false" without further qualification is not reasonable.
> the second phase (which essentially privatized Japanese state-owned lands)
This is not fully accurate. From Kim and Wang:
"The second phase of land reform, passed in 1951, redistributed most public lands—around
20% of all arable land on Taiwan (C. Chen 1961). Most of these lands were confiscated from private Japanese colonists expelled from Taiwan—in particular, the four major Japanese sugar corporations, which the Nationalists consolidated into the single Taiwan Sugar Company"
They continue
"Taiwan’s relatively strong system of land rights, intended to win the support of local landlords
for the colonial regime, made it difficult for sugar companies to amass land through coercion,
as the Dutch did on Java. Instead, the Japanese left the basic structure of smallholder farming
intact, contracting with these small farms for cane but maximizing profits by gaining control of
the milling, refining, and marketing of sugar (Cheng, Fan, and Wu 2022). The sugar companies
then used a system of advances to ensure that farmers planted sugar (Williams 1980)."
which seems to me to be a pretty close match for the standard "tenant farming reduces labor intensity" argument in favor of land reform.
Corrected the claim about the second phase — hopefully it is an immaterial misstatement.
I do think “increasing returns” with farm size can be rounded off to true.
I agree with the thesis of this essay. Does the last paragraph mean that you're a Georgist? I could be considered a Georgist, but maybe with caveats. https://zerocontradictions.net/civilization/georgism-crash-course
Edit: Never mind, it turns out that you already have an article where you talk about this.
Also, have you ever heard of Silvio Gesell and his ideas? I've been trying to get other Wikipedia editors to help me translate the German article into English, and just spread awareness about his monetary reforms in general across Wikipedia. https://en.wikipedia.org/wiki/Silvio_Gesell
I also have to hard disagree on the 'increasing returns claim.
As Amicus points out, this is one of the oldest and most entrenched debates in development economics. I have never heard anyone (except perhaps Soviet planners!) claiming that, historically, there have been 'increasing returns to scale in agriculture' The near consensus view, as of the late 1990s (see e.g. Handbook chapter by Binswanger, Feder, Deininger, 1995) was that, if anything, there were DISeconomies of scale or constant returns to scale in most crops. This explained the ubiquity and dominance of family-based and tenancy farming throughout the world for millenia. Large estates and plantations were generally viewed as arising out of distortions in policy and/or coercion. But there was a lively debate and much empirical investigation. The main issue was to disentangle was whether the commonly observed inverse relationship was due to due to contracting frictions (e.g. principal agent problems) plus technological constant returns to scale or perhaps issues of mismeasurement. Read Binswanger, Feder, Deininger 1995 for a survey of views up to that data (and World Bank policy).
It's only after the 1950s in highly developed countries such as the United States that the idea of increasing returns to scale starts to be entertained. Over time since (with the rise of large-scale farming in places like Brazil and Argentina) this had to be increasingly entertained. The World Bank's 2008 World Development Report (on Agriculture for Development) tackled these issues in depth, and for the first time acknowledges that technology had changed enough to question the old consensuses.
People cite the papers by Restuccia et al as if that had settled the debate. They bring some new modeling and evidence to the table, but their results have hardly gone unquestioned.
Ironically, the Restuccia et al papers build off Lucas' 1978 'On the size distribution of firms" paper which starts from the assumption that there are DISeconomies of scale in production, or more precisely, that there are constant returns to scale, but one factor such as managerial ability is non-traded due to contracting problems (exactly the assumption of the earlier ag literature!). So how do Adamopolous and Restuccia 2014 reach the conclusion that there are what look like economies of scale in production, and therefore 'too many small farms' in the world, supposedly imposing crippling inefficiencies? By slipping in decidedly weird assumptions about the shape of the production function (farms become less not more capital intensive as they grow in land size in ways that lower their costs). This is justified by supposedly calibrating such a function to the farm size distribution to the US economy in an exercise that lumps large grazing farms in Wyoming with wheat farms in Iowa or New England to get a farm size distribution with very large farms. They then assume that the rest of the world should follow this pattern and when, unsurprisingly, the world doesn't they declare this as evidence of huge misallocations. Gollin and Udry have paper in the 2021 JPE challenging the use of this approach and some of their more micro-studies using data from Malawi, by demonstrating that much of what these papers label as evidence of misallocation, is almost due to mismeasurement.
None of this is to say that it's not true that there very likely are "too many small farms" in the world, or that agriculture could and should become much more productive. There are plenty of reasons to believe that many farms fall below minimum efficient scale. But a U shaped AC curve is not evidence of IRS!
The debate over land reform is old and fraught. There have been many long and thoughtful takes on the subject by Alain de Janvry (lead author on that WDR report) and many others.