THE answer to the basic question is "it depends" and I will let the others describe what the issues are. But I would like to use this opportunity to discuss the fallacy of measuring bilateral trade deficits.

Consider the iPad.

According to research by Ken Kraemer at UC Irvine, the component parts of the iPad are imported to China from South Korea, Japan, Taiwan, the European Union, the US and other places for final assembly. None of the component parts are made in China: it's only role is assembly.

ONE of the big lessons we have learned for fiscal policy is that there is no such things as a shovel-ready project. So using fiscal policy for direct short-term stimulus is not helpful from a macroeconomic viewpoint. It's also bad from a microeconomic viewpoint: projects that would not pass a cost-benefit test  in normal times up being funded in the name of "stimulus".

On the other hand, "automatic stabilisers" like unemployment insurance have turned out to be pretty important. It is worth considering other such automatic stabilisers such as a German-style worksharing programme.

We have also come to a new appreciation of "automatic de-stabilisers" such as state and local government budget cutbacks. When a recession hits, state and local tax collections drop and this translates into cutbacks in state spending making aggregate demand even weaker.

I AM a huge fan of randomised trials. Last year at Google the search team ran about 6,000 experiments and implemented around 500 improvements based on those experiments. The ad side of the business had about the same number of experiments and changes. Any time you use Google you are in many treatment and control groups. The learning from those experiments is fed back into production and the system continuously improves. Wouldn't it be great if we could do that for economics!

Of course, it's a lot more difficult to do economic experiments than it is to do online experiments. But even though controlled economic experiments are costly and time-consuming, they are well worth doing since they are the gold standard for causal inference.

UNIONS have the same problem that democratic governments have: they have a tendency to sacrifice the well-being of future generations relative to current generations, since only the the current generation is able to vote.

The generous deals that the UAW negotiated in the 1950s and 1960s were great for the union members of that period. The auto companies responded by moving production to non-union states seeking lower wages and more flexible work rules, leaving the rust belt in a dire situation, so the sons and daughters of those union members did not do nearly so well.

Similarly, public employee unions have negotiated generous pension packages and work rules that are not sustainable. Those who negotiated these benefits have retired, but the next generation of public employees will not get as good a deal as their predecessors.