I'm not sure I can give you the 5-year old answer, but I can give you the health policy wonk answer.
Background: up until the early 1990s medicaid and medicare paid wildly different rates for different procedures - and also different rates to different physicians for the same procedures. As the number or procedures and tasks physicians performed increased, and the population covered by medicare expanded, the government became interest in standardizing the amount paid and reducing healthcare costs. Before this medicare would pay would they regarded as UCR (usual, customary, and reasonable).
Development of RVUs and RBRVS: in 1985, Bill Hsaio, a professor of health economics at harvard, took on the task of trying to identify how much work each procedure covered by medicare actually involved. This was a painstaking task that took 3+ years, and this work led to procedures being assigned a score called the relative value unit. This could then be multiplied by a fixed amount to give the total compensation for the procedure. This number has been falling over time, and differs between medicare and private insurance. This number is called the annual conversion factor. It is currently $37.73 for medicare. This led to the development of the resource-based relative value scale that came up with the list of all RVUs covered by medicare.
Adoption and corruption by AMA: in 1992, CMS (center for medicare and medicaid services) formally adopted this. Curiously the AMA is responsible for developing the codes (called current procedural terminology) or CPT codes for all procedures. Even more bizarrely, they are also now responsible for assigned RVUs to these CPT codes. This is done by a 31 physician committee called the "ruck" or Relative Value Scale Update Committee (RUC). Historically most of members have been specialists like cardiologists, dermatologists, gastroenterologists, orthopods (can you see a pattern here) so the number of RVUs assigned to procedures in the specialties best represented on this committee are the best paid. Conversely, primary care and psychiatry has historically been poorly represented - though this is changing. Another criticism of the RVU scale is they have not been accurately updated to take into account that many procedures (like say a colonoscopy) that took alot of time in the past but can be done in minutes now and thus require less work, which is why GI docs etc are overpaid compared to cognitive specialties where the main codes for billing are for E&M codes. Although it is ultimately up to CMS to accept or reject the RUCs recommendation they historically have accepted >90% of them
Components of the RVU: There are three components of the total RVU (or tRVU).
1) There is the work RVU or wRVU (typically about 48% of the tRVU) - for time, technical skill, effort, judgement, stress etc
2) There is the practice expense RVU or peRVU (typically abour 47% of the tRVU) - basically operational costs, non-clinical personnel, office space etc
3) There is malpractice RVU or mRVU (typically about 4% of tRVU) - estimate of relative risk of procedure
In addition there is a slight adjustment for geographic practice cost (GPCI), for example you'd get paid a bit more if you lived in Seattle or San Francisco compared to somewhere in the midwest
so total RVU = (wRVU x wGPCI) + (peRVU x peGPCI) + (mRVU x mGPCI)
and total compensation = tRVU x conversion factor (CF)
except that in 2007 congress introduced a budget neutrality adjustment factor (BNF) to scale down costs on work RVUs so:
tRVU = wRVU x 0.8994 x wGPCI) + (peRVU x peGPCI) + (mRVU x mGPCI)
MACRA: to confuse things further, from 1997 medicare physician fees were curtailed by something called the sustainable growth rate or (SGR) which unless you've been living in a cave you should have heard of as a physician since the number one priority of the AMA until last year was to lobby to repeal this. And last year, it was repealed and replaced by Medicare Access and CHIP Reauthorization Act of 2015 (MACRA). This has another way of calculating the conversion factor which we wont get into here, but essentially they are trying to correct for misvalued codes (ie codes for which physicians are being overpaid because they dont accurately reflect work done)
Fee for service: RVUs are entrenched in the fee for service model of care. Although some changes are being made to the way physicians are paid to take into account "quality measures" (though it is arguable whether these metrics are really looking at quality in a meaningful way) ultimately the ACA leaves in place the fee for service model, which whatever way you look at is is only good for physicians, but leads to an expensive system where patients get poorer quality care but more of it (more investigations, treatments etc)
For further reading see:
http://medgroup.ucsf.edu/sites/medgroup.ucsf.edu/files/the_basics_of_rvus_and_rbrvs_0.pdf
https://www.sgim.org/File Library/SGIM/Communities/Advocacy/Advocacy 101/Do-RVUs-Undervalue-Primary-Care---Primer-on-the-RUC-3-14-2014.pdf
