Not always. One of the more strident criticisms of this economic recovery (since 2008) is that the size of the workforce has shrunk because some workers are unable to reenter it after having been laid off. The increases in efficiency seemingly aren't translating into higher wages or more jobs, just more net returns to the finance folks.
So now you have unused companies that are not provided any useful service. They are mainly there for jobs for jobs sake. There will be lots of smaller companies with lots of duplicated resources, which could have otherwise have merged and shared resources.
At what point does this become silly?
You might as well pay people to dig giant holes.
If you were able merge two companies, fire people, and get rid of unused assets and still serve the same amount of customers. That means there was massive waste, and the merge resolved that.
Also we have to accept that what is long term best for "The Economy" isn't necessarily short term best for all the people in the economy.
I'm not against growth, I'm just against ignoring and refusing to pay the very real human costs of growth.
I actually think we are far from that, because human wants are virtually unlimited, so there will always be jobs required for humans in arts/entertainment/engineering. Things that will always require a human judgement.
But lets assume we are? The best way to force action is for it actually happen. If we just create non-jobs to cover it up, politicians will put off doing something about it for a long time.
But they're good for the bankers and consultants who advise on such things, and often for the incentive system the shareholders have offered the CEO. And not every acquisition destroys value, think of Google buying Youtube. Of course we are going to buck the trend, right?
[1] http://www.nber.org/digest/aug03/w9523.html [2] http://www.businessinsider.com/damodaran-the-acquisition-pro...
This generalization is justified if, and only if, there are no externalized costs and the perfect knowledge of resulting utilities pillar of the rational choice model accurately reflects reality; neither of these are true generally of real-world transactions, so the generalization is unjustified.
Like the world without friction of many introductory mechanics problems, the world of Econ 101, while it may be a useful foundation for building an understanding of the real world, is very much not the real world.