# How are Supply and Demand curves constructed?

## Question:

From my understanding of supply and demand, the price which the market sets is the intersection between the demand curve (quantity is high when price is low ) and supply curve vice versa. My question is while these are theoretically concepts, how would an economist go about forming these curves? Secondly and potentially harder to conceptualize, is when we see these curves, they are static in time however we know products such as oil and gas (focus of my questions) is moving around. Does this mean these two curves are bouncing around and not consistent "static."?

Economists always consider it important to identify whether theoretical models effectively explain the real economy. Your question has also been a long-standing one for economists. At first glance, it may seem possible to estimate a demand function with a bit of data. As in Econ 101, since the demand function (I will explain based on the demand function here, as the supply function is always the opposite) suggests an inverse relationship between quantity demanded and price, if we know the quantity demanded and the price for several transactions of a particular good, we can "estimate" the demand function as shown in the figure below.

However, can we call any curve connecting the transaction data the demand function? No. An assumption required when explaining a model in economics is "ceteris paribus," or "holding all other things constant". To estimate the demand function from those three points requires a very strong assumption that conditions other than price and quantity demanded do not change! But as we know, when the data is generated and delivered to economists, not only prices but also other demand conditions may have influenced it, or it may have changed depending on the supply factors. Since we can only observe the transaction that comes from equilibrium, the data might come from equilibrium points of different demand and supply functions as in the figure below. It would be unreasonable to use it to estimate a single demand function. Also, if we use such data, we cannot identify whether what we are estimating is the demand function or the supply function.

(demand and supply figure)

This problem has been a very important one for economists for a long time (as early as Working's 1927 paper!), and economists still consider it important to identify economic models from observable data. Going back to our example, to identify the demand function, in addition to the equilibrium price and quantity, we need factors that shift the demand function and factors that shift the supply function to control for the possibility that what we are estimating is the supply function, not the demand function.

For example, suppose we want to estimate the demand function for ice cream using data on ice cream sales prices and quantities in Iowa. This data shows that as prices increase, quantities decrease. However, there have been other important changes in the market besides prices! The excessive sugar and fat content of ice cream and its health risks were nationally reported, changing consumer tastes. On the other hand, an infectious disease broke out among dairy cattle, significantly increasing milk prices. The former is a demand-side factor, and the latter is a supply-side factor. We need to "control" for these factors in an econometric sense to understand how price changes affect consumer demand.

The second issue may seem more complex, but it can also be explained by the gap between theoretical assumptions and how data is generated from the market. In economics, we assume that demand and supply functions do not change in the short run, but this "short run" is defined as the period where only prices can change while other external factors remain constant (ceteris paribus). However, in real economy, everything can change in all directions even in a short time in the oil and gas market, which are constantly affected by global events; they are moving around. Therefore, if economists want to analyze demand in the oil and gas market, they will control for major moving factors and make assumptions so that the data can be explained by economic theory.