Keynesian economists like to talk about the paradox of thrift, a form of paradox of aggregation. In their belief, it is good for a person to be thrift, but in aggregation, saving too much means too little consumption. The economy won't grow if no one spends. I think this idea is ridiculous because there is no real person in this conjecture.
When I was studying chemical engineering in college, we were given a task to design a unit in a chemical plant. It was a crazy idea to ask college kids to do this, and, of course, I forgot everything I learned by now. Well, except one...
We were told that a reaction that works in a lab might not in a real plant. When you scale up a reaction by multiplying the amount of compounds you put in, sometimes it explodes. We were told that heat generated in the reaction is the culprit. You need to stir the reaction tank constantly to make sure heat dissipates. Once you take care of the heat problem, the same reaction happens smoothly. So there is no paradox of aggregation. We chemical engineers know what happens inside the box when we scale the model. But Keynesian economists don't know that. They watch the tank explodes and think, "uhm, things don't work in aggregate the same way as individually." Then they take that explosion as given. After decades of advance in macroeconomics, mainstream macroeconomists have already looked into the box and know what is going on (heard of price stickiness, New Keynesian models?) No matter. Keynesian economists plow on. They keep tell you government spending is the way forward. You don't need any real person making trade-offs in the economy. Government spending eases all the pain.
OK, here is the solution to the paradox of thrift. People react to incentives. If the price is right, people will buy stuff, instead of saving. When the business faces difficulties in selling to thrifts, they will lower the price. That's the price adjustment mechanism. That's the heat dissipation mechanism. There is no paradox. There are only rational people. Got it?