Evaluating the holistic costs and benefits of corn production systems in Minnesota, US

Global agriculture aims to minimize its impacts on environment and human health while maintaining its productivity. This requires a comprehensive understanding of its benefits and costs to ecosystems and society. Here, we apply a new evaluation framework developed by the Economics of Ecosystems and Biodiversity for Agriculture and Food (TEEBAgriFood) to assess key benefits and costs on the production side of genetically modified (GM) and organic corn systems in Minnesota, USA. The market value of GM corn is $4.5 billion, and only $31.8 million for organic corn using production data and market prices of 2017. GM corn generates revenue of $1488 per hectare (at $121 per MT), which is significantly lower than the organic corn at $2793 per hectare (at $294 per MT). Using a novel three-stage wellbeing valuation, analysis of the associations between corn production intensity and subjective measures of general health and wellbeing indicates that the total non-financial health cost associated with GM corn is $427.50 per hectare or $1.3 billion annually. We also find that the total annual environmental cost associated with GM corn production is $179 per hectare or $557.65 million within Minnesota. The use of the evaluation framework can help to improve decision making at farm and policy level to develop sustainable agriculture in order to minimize environmental and health related costs to society and economy.


Farming systems
The dominant corn production systems in Minnesota are summarised in Table S1.

Methodology
To measure the impact of corn production on health, we followed the well-being valuation method explained in MARCH (FAO, 2017), which offers an alternative to the Quality-Adjusted Life Years (QALYs) approach of valuing the non-financial costs of health. To use this method, we first estimated the impact of a non-market good (in this case corn intensity), income, and other determinants of wellbeing on measures of subjective well-being (SWB), such as life satisfaction. Thus, as displayed in Figure S1, the wellbeing valuation method measures two effects: the impact of the non-market good on SWB ( ) and the impact of income on SWB ( ).
Consequently, using the estimated impacts of income and the non-market good, we assessed the monetary value of the non-market good. This monetary value shows how much an individual would have to be compensated to return their wellbeing to its original level (the status quo without the non-market good).
Figure S1 Graphical representation of the WV method for valuing corn intensity as a nonmarket good.
In this study, the non-market good being valued is corn intensity in the proximity of where individuals live and the measure of SWB is their life satisfaction (In the Gallup data, life satisfaction is measured by the Cantril ladder scale which poses the following question: "Please imagine a ladder with steps numbered from zero at the bottom to ten at the top. Suppose we say that the top of the ladder represents the best possible life for you and the bottom of the ladder represents the worst possible life for you. If the top step is 10 and the bottom step is 0, on which step of the ladder do you feel you personally stand at the present time?). The monetary value obtained through the wellbeing valuation method is the well-being effect of corn intensity through its impacts on health. This monetary value can also be interpreted as the (non-financial) costs of health associated with corn intensity. Note that the WV method does not account for any health impact caused by the consumption of products containing corn (as corn intensity can be assumed to be independent of the amount of corn consumed, not including corn consumption in the model does not bias our estimates).

Variables
We used following variables suggested in Fujiwara and Campbell (2011)

Monetary value
The valuation of non-financial health costs of corn production is based on the well-being valuation method. First, we estimate the impact of corn intensity (δ) on general health.
From Table S2, estimating the model using 5 km and 10 km buffers results in a value for δ of -0.0021 and -0.0025, respectively (statistically significant at the 5% level). The sign of δ indicates a negative association between experienced corn intensity and an individual's health. Based on these estimations, the negative impact is statistically significantly higher in the 10 km buffer compared with 5 km buffer. As the geographical area of a ZIP code is large (The average, minimum, and maximum values for the ZIP code geographic areas in Minnesota are 244 km 2 , 0.15 km 2 , and 289 km 2 ), particularly in rural areas, the corn intensity measured in a 10 km buffer more precisely represents the intensity experienced by individuals throughout the ZIP code area.
The value of δ implies that an increase in corn intensity in a 10 km buffer by 1% will decrease general health by 0.0025 points (where general health is measured by a 1-5 scale). Increasing the corn intensity by 9.74%, which is the average land used for corn production in 10 km buffer in our sample, implies a decrease of 0.024 points in general health. In other words, going from no corn production to the average level of corn production (holding other factors in the model constant) implies a 0.67% decrease in general health (relative to average levels of general health).
Consequently, we apply the wellbeing valuation method to estimate the non-financial health costs of corn intensity in the respondents' surrounding area. The non-financial health costs associated with a 1% increase in corn intensity in the vicinity of an individual's residence is $20.7 per year in the 5km buffer and $24.7 per year in the 10 km buffer. These results are based on Minnesota average household income 2016 which, according to US Census Bureau, was $83,100. Our analysis in this part reveals that although 26% of the sample have at least one organic farm with some corn production in their ZIP code, organic corn farms are particularly small on average comprising 0.3% of total land used for corn production. This means that the proportion of sample likely to be impacted on by an organic corn farm is much lower than that of a non-organic farm. The relatively lower size of organic corn farms might be an important issue in identifying the health impacts of organic versus non-organic corn production in case that we have access to a more complete and detailed data set.

Aggregating Health Costs:
To calculate annual non-financial health costs associated with corn production in Minnesota, we follow the following steps: 1. Using the wellbeing valuation method, we obtain the monetary value of the average health costs on individuals of a 1% increase in the intensity of corn production in their respective ZIP code. 2. Using data from the United States Census Bureau, we find the population for each county in Minnesota. The population of each county is then multiplied by the health costs obtained in step 1. This will give us the health costs per 1% corn intensity in each county. Note that our estimates are based on a sample of respondents aged over 18. If we assume that individuals aged under 18 are not affected differently by corn intensity, we multiply the whole population of each county to the health costs per individual. For example, for Dakota county, with a total population of 421,751, the annual health costs associated with a 1% increase in land used for corn in a 10 km buffer will be: 421,751 × $24.7 = $ 10.42 million. 3. To calculate the health costs of corn production per county, we multiply the number obtained in step 2 by average corn intensity for the county. For example, In Dakota county, the average corn intensity in a 10 km buffer is 8.63%, so the annual health costs of corn production are: 10.42 × 8.63 = $ 89.90 million a. Table S3 shows 40 counties with the highest health costs of corn production based on relative intensity in a 10 km buffer. 4. Finally, we find the total health costs in Minnesota by aggregating all counties health costs. Based on our model, the annual non-financial health costs of corn production in Minnesota are about $ 1.3 billion (approximately $233 per capita). This is broadly aligned with the costs obtained in MARCH (FAO, 2017). In this study, the annual nonfinancial health costs of the UK food system for different health problems were between $107 per capita and $1372 per capita.