I often see claims that new (generally government) projects will "create n jobs".

For example, this 2013 claim from the Economic Development Blog about the 2020 Olympic Games:

The Olympic Games are expected to create 152,000 jobs nationwide across Japan. This includes 84,000 new jobs in Tokyo and 68,000 in the rest of the country.

I have always found such claims to be vapid, because it doesn't say how long the jobs last.

Sticking to the example: One of those new positions might be to be a security guard for a month during the games. One of those new positions might be to work on the planning for the event for seven years. One of those new positions might be to be a groundskeeper at a new Olympic facility for decades.

Lumping these together seems like it is a misleading metric. Perhaps saying "250,000 job-months" would be a better measure?

But I am economically naive. Perhaps economists have already considered this and have a standard definition I don't know about to allow them to properly compare.

Are these figures mostly meaningless, just being used to hype a project by its proponents? Are they accepted estimates of real economic boost a project might bring?


1 Answer 1


There is no standard, internationally agreed definition of what should count as "a job" when conducting economic impact assessments. However, it is an important question that is often (not always) taken seriously when constructing such estimates.

Sometimes it is very important that jobs estimates are comparable. This is particularly the case when politicians have to decide between competing investments. If project A creates 1000 jobs and B creates 1200 jobs, then you should be sure that those estimates used the same methodology (and that it was sane) before deciding between the two. For this reason governments often have official (though maybe internal) guidance on how to calculate the number of jobs created. This is particularly true in the international development sector, where it is now quite common for government interventions to be aimed primarily at job creation.

On the other hand, it is not necessarily the case that we would want to use the same definition in all contexts, even if it's the same government creating estimates. For instance, the UK government may want to use different criteria for assessing a community regeneration project in northern England than it would for an aid project in Afghanistan. It would hopefully be transparent in both cases.

Each case is complicated for lots of reasons, including those you mention and many others. How many hours a week must you work to have a job? Does that change depending on what local working patterns are? What if your job is seasonal? Does it make a difference if you left another job to take the new one? Does it matter if your previous job was in the informal sector, or part time, or not as well paid? Does it matter if your job conforms with the ILO core conventions? What about the new jobs that would have been created anyway? And what about indirect jobs created as a result of increased spending of newly employed people (the multiplier), or because people copy you (spillovers)?

EDIT: There is a comparison of some common job creation definitions for development in a DCED working paper on pp12-13.

  • 2
    $\begingroup$ Thank you. My take-away is that while there may be a systematic way of generating this challenging figure, a number quoted in a newspaper without the context to indicate the method used is largely meaningless. $\endgroup$ Aug 29, 2018 at 2:33
  • $\begingroup$ Yes, I agree. It's a case in which press articles should always reference an accessible source, at least online where that is easy. Since it is actually quite challenging to estimate job impacts well, a half-intelligent journalist should have a pretty good chance of finding questionable logic in the estimate if they assess critically. $\endgroup$
    – Dan
    Aug 31, 2018 at 18:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.