Belief: Rational agents make optimal decisions and firms maximise profits, but the economy is susceptible to real shocks like unanticipated policy changes, technology advances and changes the supply of raw materials.
Policy recommendation: The government should take a laissez-faire approach and leave the economy to work by itself.
New Classical economics was developed by Robert Lucas Jr. as a response to the failure of Keynesian economics to explain the breakdown of the Phillips curve in the 1970s. This school of thought revolutionised economic thinking as it attempted to explain macro phenomena with micro concepts.
New Classical economics builds macro models based on micro foundations; for instance, it assumes that agents have rational expectations, prices and wages are flexible, there is always full employment and markets always clear.
One of the major propositions by New Classical economics is the theory of rational expectations. This posits that agents will look at all available relevant information when forming their expiration.
Using their models, New Classical economists argue that any announced policies will be incorporated by agents into their rational expectations and will, therefore, have no real effects, only nominal. For instance, if the government announce that they will increase government spending then agents will recognise this and adjust their behaviour accordingly. Firms will increase their prices and workers will demand higher nominal wages. The rise in government spending will only lead to a change in prices and wages, not production.
The only way the government can affect real variables, like output, is by using unannounced policy intervention, because agents cannot incorporate these into their expectations as they do not know about it. However, unannounced policies will only take effect in the short-run as the economy will adjust back to its long-run equilibrium after agents realise that the government fooled them with an unannounced policy.
New Classical economics met with a nasty fate in the empirical realm, however, as it turned out that the predictive and explanatory power of such models was very low.