Do we produce enough food already?

Earlier this week I had the pleasure of giving the George Morris AgriFood Policy lecture at the University of Guelph.  I primarily focused my talk on the benefits of food and agricultural technologies and the importance of productivity growth for solving our future world food problems.  

At the conclusion of my talk, an audience member played devils advocate asked an important question that deserves more widespread discussion.  In short, the question was something along the lines of the following: don't we produce enough food already?  It is a question reflected in many popular writings.  This headline, for example, is "We Don't Need to Double World Food Production by 2050." Here's Mark Bittman writing in the New York Times: "The world has long produced enough calories . . .".  Here's Bittman again under a headline in the same outlet "Don't Ask How to Feed the 9 Billion" because, in his words, "The solution to malnourishment isn’t to produce more food." 

Here are my main main thoughts on this line of thinking:

1) Even if we produce enough calories today to meet today's population, that doesn't mean we produce enough for tomorrow's population.  Productivity growth is gradual and incremental, and if we found ourselves in a situation of needing more food, the new technologies to produce them cannot be created over night.  This is particularly true of our ability to produce in the future is hampered by climate change.

2) There is no binary category of "enough food."  Greater food production leads to lower food prices and lower food insecurity.  I haven't yet met a food consumer who wouldn't prefer paying lower food prices, holding quality constant. 

3) I may be true in an accounting sense that we produce enough calories today to meet total caloric needs.  But accounting isn't economics, and we need to consider the incentives of the system that produces the sufficient calories today relative to an alternative system that is either less productive or involves widespread redistribution.  Massive redistribution of food can destroy the incentives of people to produce the food.  One cannot disentangle the fantastic productivity of our current system with the market forces that led to it's origin.  Stated differently, there is no reason to imagine we'd produce the same number of calories if "the system" were changed to one with massive confiscation/redistribution.  Brady Deaton altered me to this fascinating paper in the Journal of Political Economy showing that 75% of the increase in China's agricultural productivity after 1978 was due to strengthening of individual incentives.

4) It's important to look at productivity through the lens of sustainability.  Higher productivity means getting more (or the same) amount of food output using fewer inputs and resources.  Are people really wanting to argue that they'd prefer systems that require more of our natural resources - more land, more water, more fossil fuels? Since when is lower productivity and inefficiency preferred?  Even if "enough" food is produced today, improved productivity means we can keep producing the same quantity but shrink agricultural's footprint on the land, use less water, fewer pesticides, etc.    

5) If the solution to the food problem is simply shipping food from high productivity countries and sending (or stated more pejoratively "dumping") in lower productivity countries with hungrier citizens, this may harm the livelihoods of producers in low productivity countries and reduce their incentives to adopt efficient forms of agriculture.

6) If places like the US decided to forego new food and agricultural technologies and farmers were forced or incentivized to adopt lower productivity systems, what would happen to patterns of global trade and production.  US farmers compete with farmers all over the world to serve US consumers and consumers worldwide.  Not only would such policies likely reduce US exports, it would make imports relatively more attractive.  Is the solution then import tariffs to prop up our lower productivity system?

7) One can go back to writings from over 100 years ago and find claims that the problem of production and scarcity had essentially been solved, and all that was needed was a heavier handed state to ensure "fair" distribution (e.g., see Edward Bellamy's Looking Backward, published in 1888).  Imagine the world we would live in today if that premise were widely accepted back in 1888 - that the state of production was "good enough" and we could stop worrying about growth and progress.  How much growth would we have lost out on had we stopping innovation in 1888?  We'd still be hand-picking cotton, planting with mules, eating much more salt- and vinegar-cured meats, and more.  What will the food and agriculture future look like in 2088, and what will we give up if we stop working on productivity-enhancing technologies today? 

Agricultural Productivity and Food Security

There are two competing narratives about the future of food.  One is that the world population is growing and we need to increase agricultural productivity to "feed the world".  The other argument is that we don't need to produce more food - we already produce enough food to feed the world and our problems are really more about distribution than production.  Folks in the later camp often advocate for lower-productivity forms of agriculture that they perceive to have health or environmental benefits.  Like most arguments, there are elements of truth to both sides.  

As a proponent of improved agricultural productivity (which, I've argued is the key metric to improved sustainability), it bears asking: if a country's agriculture is more productive are it's people better fed?  

To delve into this question, I combined two data sets.  The first is a measure of a country's agricultural productivity from the World Bank in the year 2015.  In particular, they calculate for a large number of countries, the agricultural value added per worker.  In their words:

“Agriculture value added per worker is a measure of agricultural productivity. Value added in agriculture measures the output of the agricultural sector (ISIC divisions 1-5) less the value of intermediate inputs. Agriculture comprises value added from forestry, hunting, and fishing as well as cultivation of crops and livestock production. Data are in constant 2010 U.S. dollars.”

By this measure, the most productive countries are Slovenia, Singapore, Norway, France, Lebanon, Canada, New Zealand, Finland, and the United States, each of which produced more than $80,000 in agricultural value per worker in 2015 (measured in 2010 dollars).  Places like Malawi, Congo, Mozambique, Gambia, and Madagascar had some of the lowest productivity, with agricultural value added at around $400/worker or less.

Secondly, I collected data from the Global Food Security Index, a project ran by The Economist and supported by DuPont.  In their words:

“The Global Food Security Index considers the core issues of affordability, availability, and quality across a set of 113 countries. The index is a dynamic quantitative and qualitative benchmarking model, constructed from 28 unique indicators, that measures these drivers of food security across both developing and developed countries.

This index is the first to examine food security comprehensively across the three internationally established dimensions. Moreover, the study looks beyond hunger to the underlying factors affecting food insecurity. This year the GFSI includes an adjustment factor on natural resources and resilience. This new category assesses a country’s exposure to the impacts of a changing climate; its susceptibility to natural resource risks; and how the country is adapting to these risks.”

From this project, I pulled each country's overall food insecurity score (calculated in September 2017), which took on the values of around 30 for countries like the Congo, Madagascar, Chad, and Malawi, and was above 80 for countries like the U.S., the U.K., Ireland, and France.  Although they call this a measure of food insecurity, a higher score actually means a country is more food secure.   

So, what did I find?

There is a strong positive relationship between a country's agricultural productivity and how well it's people are fed and how food secure they are.  Fitting a logarithmic relationship between the two variables suggests that 82% of the variation in the food security scores across countries is explained by differences in agricultural productivity.  

Now, there are a lot of other things going on here as agricultural productivity is likely correlated with and affected by other factors affecting a country's general productivity and development, but the above figure might give pause to those arguing for lower productivity forms of agriculture. 

At the top end, the curve suggests one can sacrifice some productivity with only a small reduction in food security (going from $80,000/worker to $40,000/worker) reduces the food security scale from about 80 to 75.  But, at the lower end, going from, say, $20,000 in agricultural output per worker to $10,000/worker reduces the food security scale from about 70 to 60, and reducing productivity another $10,000 lowers the food security scale down to the 30s.  

Technology and evolving supply chains in the beef and pork industries

That's the title of a new article in Food Policy written by Josh Maples, Darrell Peel, and me.  The paper will ultimately be part of a special issue on technology and supply chains. 

Here is part of the lead in.

“The structural shifts in the beef and pork industries have occurred alongside (and perhaps because of) technological innovation and its effect on the flow of these meats through their respective supply chains. Technology innovations have been a major factor in the changing economics of the beef and pork industries. Improved nutrition, growth promotion technologies, better genetics, and economic conditions have all played a role in livestock becoming more efficient (Lusk, 2013). The values of improved technologies between 1977 and 2012 in the beef and pork sectors have been estimated at $11 billion annually and $7.6 billion annually, respectively (Lusk, 2013).”

We discuss the nature and causes of different market structure in the beef and pork industries.

“Even with the presence of marketing agreements, the beef industry is easily the least vertically integrated of the big three protein industries (Ward, 1997). The key reasons for this revolve around the aforementioned asset specificity as well as the biological makeup of cattle. There is a greater incentive to vertically integrate or engage in contracting in livestock industries in which genetic changes can be made more rapidly (Ward, 1997). The biological production cycle is about two years for cattle, which is twice as long as that of hogs and the genetic base of cattle is relatively diverse and is not narrowing (Ward, 1997). Alternatively, market coordination has allowed the pork industry genetic base to narrow toward the most efficient hogs for production. The number of hogs marketed today is 29 percent greater than in 1959 from a breeding stock that is 39 percent smaller (Boyd and Cady, 2012). Geographical concentration also plays an important role. During the cow-calf stage, cattle are scattered throughout the U.S. due to the required land and forage needed while hog production is centered in the Midwest (and more recently the Southeast) near the heaviest corn-producing states. These factors create significant barriers to integration in the beef industry.”

And, we discuss the impacts of various technologies on the industries.  Here's a segment on effects of pharmaceutical innovations in the cattle industry.

“Vaccinations, parasite control, ionophores, antibiotics, growth promotant implants (often referred to as growth-promoting hormones), and beta-agonists have been the most widely-used of these innovations (Arita et al., 2014 ; APHIS, 2013). The productivity and economic impacts of these technologies are large. Lawrence and Ibarburu (2007) estimated that the cumulative direct cost savings of the technologies was over $360 per head for cattle over the lifetime of an animal while Capper and Hayes (2012) estimated that the increased cost of U.S. beef production without growth enhancing technologies would be the equivalent of an 8.2 percent tax on beef. Elam and Preston (2004) discussed each of these technologies at length in their summary of the technological impact in the beef industry. They found that growth implants increase rate of gain by 15–20 percent and improve feed efficiency 8–12 percent. Growth-promoting hormone implants are believed to be used on approximately 90 percent of cattle in U.S. feedlots (Johnson, 2015). Elam and Preston (2004) also found ionophores increase average daily gain by 1–6 percent and improve feed efficiency by 6–8 percent. Lawrence and Ibarburu (2007) used a meta-analysis approach to find estimates for the farm level economic value of these five technologies in the beef industry. They estimated that beta agonists improve feedlot average daily gain by 14 percent and that the combination of implants, ionophores, antibiotics and beta-agonists account for a 37 percent increase in average daily gain in feedlots. These increases in feed and gain efficiency have direct effects on the profitability per animal. Lawrence and Ibarburu (2007) estimated that sub-therapeutic antibiotics impact cattle profitability by $5.86 per head, ionophores have an $11–$13 impact, and the use beta-agonists impacts per head profitability by $13.02 per head. The use of growth promoting implants has the largest impact on cattle profitability at between $68 and $77 per head ( Lawrence and Ibarburu, 2007; Wileman et al., 2009).”

National Academies Town Hall

Last week I gave a short talk at a Town Hall held at the National Academy of Science Building in Washington, D.C. in relation to the Science Breakthroughs 2030 project aimed at identifying strategies for food and agricultural research.  

You can see all the presentations here.  Or, if you just want to see my comments and provocations entitled "Importance of Understanding Behavioral Responses to Food and Health Policies", the video is embedded below.

How Expenses Vary with Farm Size

I've been a bit surprised at the number of comments and questions I continue to receive about this article I wrote for the New York Times almost a year ago.

Here are the opening sentences from the piece:

“There is much to like about small, local farms and their influence on what we eat. But if we are to sustainably deal with problems presented by population growth and climate change, we need to look to the farmers who grow a majority of the country’s food and fiber.

Large farmers — who are responsible for 80 percent of the food sales in the United States, though they make up fewer than 8 percent of all farms, according to 2012 data from the Department of Agriculture — are among the most progressive, technologically savvy growers on the planet. Their technology has helped make them far gentler on the environment than at any time in history. And a new wave of innovation makes them more sustainable still.”

Common questions I tend to get are "who are these large farms" and "do large farms use more or less fertilizer or chemicals than small farms?"  On the first question, I simply rely on USDA's classification of farms based on gross sales (which is where the above 80% from 8% originates).  The second types of questions are much more difficult to answer as there isn't great data easily accessible on the matter.

However, I recently ran across this USDA, National Agricultural Statistics Service (NASS) publication that reports farm expenses for different sized farms (again, where size is determined by gross sales).  These data are part of the Economic Research Service (ERS), Agricultural Resource Management Survey (ARMS).  Using the 2016 data in this publications, I created the following charts to help provide some perspective on how relatively small, medium, and large farms allocate their spending.

Here are relatively small farms.

The spending of relatively medium-sized farms is illustrated below.

Finally, here are graphics on spending by the largest farms.

A few comments on the comparisons are in order.  First, as indicated by the share of spending on livestock, poultry, and feed, there are different types of farms across size categories, so it's a bit like comparing apples to oranges.  The largest farms are most likely dairies, feedlots, or hog/poultry operations.  The proportion of crop output (as a share of total output) is likely higher for small and medium sized farms.  What we'd like to compare are small crop farms to large crop farms, but that data wasn't easily obtainable.   

The figures show that three categories of spending (as a share of total spending) fall as farms sizes increase: farm improvement and construction, tractors and trucks, and taxes and interest. This relates to some of what I argued in the NYT piece:  

“But increased size has advantages, especially better opportunities to invest in new technologies and to benefit from economies of scale. Buying a $400,000 combine that gives farmers detailed information on the variations in crop yield in different parts of the field would never pay on just five acres of land; at 5,000 acres, it is a different story.”

On two of the issues which people worry about the most - chemicals and fertilizers - these expenses tend to increase (again, as a share of total expenses) as size goes from small to medium than falls when going from medium to large.  However, some of this change is almost certainly due to the different mix of crops vs. livestock in the different size categories, so it's difficult to draw much of a conclusion from these data.   

Finally, I'll note the small sized categories of farm (less than $10,000 in gross sales) lose money on average.  Why?  Because, by definition, they're  bringing in less than $10,000 in revenue, but they're spending $13,755.  These farms need to generate at least $3,755 in additional annual value per farm to the farm owners, to their patrons, or to their neighbors that isn't reflected in market price for their activities to yield a net benefit to society.