Measuring unemployment is actually very difficult, and when you mix in the general political motivation for incumbent governments to report unemployment to be as low as possible and opposition groups to report it to be as high as possible, it becomes almost impossible to get a true figure.
For example, there are invariably different forms of structural unemployment in an economy, and the size of the actual population who are able and willing to work is not the same as the total size of the population. For instance, governments don't usually count children or people who have retired as unemployed, even though they are people who are not in work. Similarly some people are incapable of working due to disability. Some people choose not to work because they are in a relationship where they are financially supported or they have savings/investments enough to support themselves. Some people are unemployed because they are lazy and do not want to work. Some people have the wrong skills and/or qualifications for available jobs. Some people see certain jobs as "beneath them" and will refuse to work below a certain pay grade, or may not be willing to take a step down when in six months time another more appropriate and highly-paid job might come along. Some people are forced to leave work to be full-time carers or parents. Some people are short-term unemployed in very fluid jobs markets and may only spend a few days or weeks without a job.
Considering how complex this is, it is tempting (and many governments in fact do) measure unemployment by the "claimant rate", which is equal to the number of people currently claiming unemployment benefits of some kind. However, this measure too fails to account for a number of reasons why people are not working, so must be compared to estimates on the size of a country's available labour, and fails to account for reasons that a person may not be working but may be claiming, or claiming whilst working, and so forth. There are various other measures that are similar to this, and most rely on similarly complex reporting methods.
These figures are very easy to manipulate on both the incumbent and the opposition sides. It's very easy to point to a figure and say "they're lying, this is the true unemployment rate" without context, and since the issue is extraordinarily complex it is very easy to mislead people. Essentially understanding what particular employment rates mean and why they are at the levels they are is something that needs serious analysis, and is unlikely to be the sort of thing you can use as a soundbite whilst retaining any sort of accuracy or context.
