The article validates economic models by comparing them to real data. The researchers use a method that looks at how certain variables react to economic shocks to see if the model matches reality. They find that two types of models don't fully explain the behavior of key economic variables. Even when they tweak one of the models, it still doesn't match the data well.