This course takes a quantitative, hands-on
approach to understanding the challenges of
feeding the human population of the planet
Earth. We’ll discuss topics of nutrition,
subsistence food consumption, and consumer
demand for food to develop our
understanding of the current situation.
We’ll then develop both theories and
computer models of population dynamics
taking into account people’s decisions
about child-bearing, changes in mortality,
and changes in food supply in order to
learn something about the future of food.
Focus throughout the course will be on
developing practical tools to work with
real-world data. Those tools will include
linear programs, globally regular demand
systems, and a variety of
econometric tools. The course will rely on
a knowledge of the programming language
python.
Data8, EEP100 or equivalents required; Math 54 recommended.
Meet two times per week; Tuesdays and Thursdays 2pm–4pm (Pacific Time).
We’ll have a mix of lectures and discussion (typically on Tuesdays) and tutorials and group work (typically on Thursdays).
We’ll meet in person in Social Sciences 110. Office hours may be in person or virtual.
- Instructor: Ethan Ligon (ligon@berkeley.edu)
- GSI: Scott Kjorlien (scottkjorlien@berkeley.edu)
-
Questions/Discussion: Please use https://edstem.org to ask questions of the instructor or assistants; we’re likely to miss email. Note that you can post anonymously and/or privately.
-
Instructor’s Office Hours: You may also make an appointment to speak with Professor Ligon during his office hours via https://are.berkeley.edu/~ligon/appointment.html.
-
Office hours (GSI): Scott will hold an office hour 11:00am on Tuesdays in Giannini 203, or by appointment.
Grading in course depends primarily on peer evaluation. In particular, every project will lead to you evaluating your classmates in two main ways, and being evaluated yourself in three.
Everyone will evaluate every team according to several criteria. For example,
-
Code
- Did code work as intended?
- How elegant was the team’s code?
- How ambitious were design goals?
- How completely were design goals met?
-
Organization
In the presentation:
- How well did team manage its time?
- How well did the team work together?
-
Style
- How interesting was the presentation?
- How polished was the presentation?
-
Overall
- If you were in the position of needing to hire a team to do this sort of analysis, would you hire this team?
- Other constructive comments & criticisms.
- Finally, we’ll ask you to rank all teams according to your overall
impression of their presentation.
Everyone will also evaluate all of the individuals on their team according to several criteria:
- Quality of work
- Could be counted on to complete tasks in a timely fashion?
- Helpful to others in group?
- Contributed to the smooth working of the team?
In addition, we’ll ask you to give some additional information about each of your teammates, indicating:
- What were each person’s main strengths?
- Would you like to work with this person again?
And finally, we’ll ask you to:
- Rank each person according to their overall contribution to the project.
Your own evaluations are an important individual contribution to the class, and the quality of your evaluations will affect your grade. There are three criteria we’ll use in judging the quality of your evaluations.
-
Prediction of others’ evaluations of you
You’ll provide evaluations not only of other teams, but also of your own team. And you’ll evaluate not just your teammates’ contribution to the project, but also your own contribution.
Your self-evaluations will affect your grade. However, the way in which these will affect your grade will depend not on how good you say you are, but how accurately you predict how others evaluate you. In particular, the closer your guesses about others’ evaluations are to the average of what others give you the higher your grade.
-
Information in your own evaluations of others
The greater the information provided by your evaluations of others the higher your grade. The amount of information will be measured partly according to the variation of your evaluation of others, and partly according to a (subjective) measure of the quality of your comments.
Observation: if you give everyone the same scores (e.g., everyone gets top score) there is no variation in your evaluation. This would negatively affect your own grade.
-
Correlation with evaluations of others
Your evaluations must be honest, in the sense that they are attempts to fairly evaluate the efforts of others and of your own efforts. Ideally there will be broad agreement across different people’s evaluations. If your evaluations are `outliers’ then this may negatively effect your grade. Further, if upon examination it appears that you’ve used your evaluations strategically there may be further repercussions, most particularly if the manner in which you’ve evaluated others violates Berkeley’s Honor Code (i.e., you must “act with integrity, honesty, and respect for others”).
For each project, you will also be evaluated in terms of your engagement. This is the one category which will be evaluated solely by instructors and GSIs, rather than by your peers. Particular things we will look at:
- Number and quality of comments, etc. on edstem.org.
- Number and quality of comments and issues on github repositories (both yours and others).
- Number and quality of comments or PRs specifically pertaining to issues/bugs in code or material in EEP153 repos, including
The course revolves around a sequence of topics, each exploring a substantive issue involved in “feeding the planet” and each introducing novel tools. Students will work in small groups to complete one structured project for each topic.
Students will review introductory materials about coding (Python and Pandas) and potential ways to collaborate (Git, github, Google Colab, Trello, and Agile). See online posts for links to resources.
Students will construct datasets on the distribution of characteristics in the world population, including measures of resources, and the age and sex composition of the world population. A separate dataset allows us to think about food supply.
- Malthus An Essay of the Principle of Population (1798)
- de Janvry-Sadoulet (2015), Chapter 11 of Development Economics
- Fuglie (2012), https://www.ers.usda.gov/amber-waves/2012/september/global-agriculture/
- NYTimes “What to eat on a burning planet” series, especially:
Every living human has some minimal, or subsistence, nutritional requirements; should these not be satisfied health and even life may be threatened. People satisfy these needs by eating various kinds of food, but there may be many different food diets which satisfy people’s subsistence requirements. One criterion for choosing among these diets is cost.
In this topic students use contemporary data on different kinds of foods available to the US population along with prices to construct estimates of minimum cost subsistence diets.
In practice, even very poor people seldom choose their diets on the basis of minimum costs. Instead, people balance nutritional requirements against considerations of cost and what we might call the gastronomical value of different diets. Here we explore the theory of demand as it pertains to these diets—how does demand for food depend on income, prices, and other observables? How well (or poorly) do these diets serve nutritional ends?
- Review basic demand theory (e.g., Chapters 3–5 in Nicholson-Snyder)
Students will use data on household food expenditures for populations from different countries to estimate systems of food demand, and relate these demands to the subsistence diets calculated in the earlier topic.
With these results in hand you will construct aggregate demand functions that allow one to make predictions regarding how aggregate demand for different kinds of foods depends on the distribution of resources and the demographic composition of the global population.
This project exploits our work on demand for food and is focused on evaluating what kinds of policies might be effective at improving nutritional outcomes for particular populations. Our earlier work addressed the question of how demand for different kinds of food depends on prices, budgets, and household characteristics, taking as given prices, budgets, and so on.
One of the take-aways from our earlier project is that the food people choose to eat may be quite different from the foods that people should eat, from a nutritional perspective.
But if dietary choices respond to prices and budgets, it may be possible to manipulate nutritional outcomes by changing either prices or budgets. We can assess the costs of this kind of manipulation (e.g., the deadweight cost of a tax or subsidy); where these costs are large we can also think about the value of innovation in either the desirability or nutritional content of food.
- Technical change: Borlaug (2000), Ars Technica (2019)
- Changes in budget: Deaton-Dreze (2009)
- Changes in relative prices: Falbe et al (2016)