Note: This is a question on the didactics of microeconomics, directed to those of you who have some experience teaching this subject.
When I studied the basic principles of microeconomics, a price-setting monopoly made perfect sense to me, but price-taking firms under perfect competition never did. No single firm can influence the market price individually, but collectively they do? Nobody sets a price for their product, but magically some "market price" falls from heaven? That seemed like a clear contradiction to me.
Then I learned about Cournot competition. Again, firms competed in quantities and the market price fell from heaven via the indirect demand function. Again that didn't make any sense to me. Real firms don't just produce some quantity and then "throw it on the market", whatever that means, to watch the market price magically materialize out of nowhere. Demand is a causal consequence of price, not the other way round. For me, in the real world, all firms were always competing in prices. So why study Cournot at all?
Now that I am teaching microeconomics to undergraduates myself, I still struggle with the question of how to motivate and explain these basic models. Should I keep things simple and insist on just assuming quantity competition under Cournot, and additionally price-taking behavior under perfect competition, thereby risking to lose the connection to the real world and establishing a kind of shut-up-and-calculate culture (which might actually make sense for quantum physics, where it comes from, but hardly for introductory economics)?
Or should I start with a more realistic approach of price competition, talk about capacity constraints and capacity-then-price competition a la Kreps and Scheinkman (1983) to justify Cournot? And then explain that the limit case of Cournot with the number of firms going to infinity approaches a perfectly competitive market under some mild assumptions, thereby having to reverse the usual order of introduction of these models and also risking to overtax my students?
I have variously tried both approaches, and mixtures thereof, but I have never been fully satisfied in the end. What is your approach and/or recommendation?
I'm not so much interested in comparing model outcomes with the real economy, but more with capturing the behavior of agents within the models, in a sense with the "microfoundations of introductory microeconomics".