In my view, it's clear that MMT as a theoretical framework encapsulates both descriptive and prescriptive elements. The descriptive part of MMT aims to describe the functioning of a modern monetary system and shed light on the nature of government finance. The main emphasis here is to explain that modern currencies cannot be understood without considering the legal and institutional context within which a modern, sovereign government (i.e., that has control over its own currency) exercises its authority. Specifically, the government's power to levy taxes and declare what it will accept in payment of taxes are the two important components for understanding modern currencies and identifying the options available to the government for financing expenditures and setting prices (e.g., of interest, manpower, commodities, etc). Flowing from this understanding of government finance is the fact that modern governments do not face operational financial constraints, but rather face self-imposed, legal ones that result from past political decisions, as well as constraints imposed through convention. Balancing the government budget on a yearly basis is one example of a self-imposed constraint.
As for the prescriptive part of MMT, my understanding is that it focuses mainly on (1) the implementation of a job guarantee (aka, buffer stock employment model) or an employer of last resort scheme as a way to achieve full employment and price stability, and (2) the application of a monetary policy where the central bank sets the nominal rate to zero, thereby letting the real rate adjust endogenously (resulting, in most cases, to a negative rate due to inflation). From a policy standpoint, I don't have any problem with the former, as I believe that such a proposal could gain considerable traction once the public understands how the policy would function. However, I can understand why some people have reservations regarding the viability of the proposal to put forth a zero nominal interest rate, as it would most likely lead to a negative real rate of interest. A good alternative, in my view, would be for the monetary authorities to set, as much as possible, the real interest rate equal to the rate of growth of labour productivity. In this way, the real rate of interest should remain mostly positive and an amount of money equivalent to one hour of labour time, if lent at that rate of interest, would continue to be worth one hour of labour time when paid back with interest.
But, setting aside this very minor point of divergence, it should be pointed out that MMT falls under the larger body of thought known as 'chartalism' or 'modern money approach', which is not tied to any specific policy prescription. This means that it isn't really a problem if people have different views about the prescriptive part of MMT: the chartalist or modern money approach allows for a wider range of points of view and policy recommendations. In a good article on chartalism, modern money economist, Pavlina Tcherneva, explained this point as follows:
...it is important to point out that Chartalist propositions are not necessarily tied to any particular policy prescription; they are simply a way of understanding the state’s powers and liabilities and its financing and pricing options. (2006:81)So, to conclude, if you find MMT's framework for understanding the implications of modern money useful for analyzing the economy and government policy but aren't sure about the policy proposals associated with the framework, there's no big problem: you're still applying the modern money approach.
Rochon, L-P, and M. Setterfield (2008), "The political economy of interest rate setting, inflation and income distribution", International Economic Policy Institute, Working paper series, 2008-01
Tcherneva, P. (2006), "Chartalism and the tax-driven approach to money", in P. Arestis and M. Sawyer (eds), Handbook of Alternative Monetary Economics, (Northampton, MA: Edward Elgar), pp. 69-86.