Big infrastructure projects have the problem that they cost a lot of money and time. Another disadvantage is the inherent problem all Keynesian ideas have, they very much depend on timing. If you start such a project too late you severely harm a recovering economy, start it too early and it won't help at all or even have a negative impact.
Also, there is no way to show whether a government project is profitable or not. This means we can't tell whether the project actually helped or hurt the economy.
Infrastructure projects don't have to be profitable that is neither the goal nor does anyone care. The goal is to create a lot of jobs and a lot of income relatively fast to increase demand which helps the economy.
The most popular types of jobs programs involve state tax breaks or subsidies that seek to seduce a company from one state to another. While this can mean good news for “business-friendly” states like Texas, such policies don’t add to overall employment so much as they just shuffle jobs around.
Such subsidies can also seduce people from one country to another. And in that sense, they would indeed create jobs.
The same argument applies to the industry-specific legislation he mentions later.
Finally, he doesn't mention, on the first page, the very straightforward way that politicians can create jobs: by spending money on big infrastructure projects. Those certainly do create jobs, and they have been a common method to do so in a recession.
To top it off, lumping politicians in with "everybody else" seems unfair. As argued above, even politicians can and do "create jobs", but it is ludicrous to argue that the founders of Google or eBay or Paypal didn't create jobs. They created tens of millions of jobs that people live off of, in the US and in other parts of the world.