What criteria should guild when to cut some program’s budget? We must first get beyond the fact the any cut will result in having less of something. If it is inefficiency or corruption that we give up—good riddance. But usually, it will be something that has some value. That does not necessarily mean that the cut should not be made.
Consider this example from my in-tray today:
“The Trump administration has made drastic cuts to the National Oceanic and Atmospheric Administration (NOAA) that threaten to impact weather forecasting and other key services provided by the agency.
In the wake of the wave of dismissals this week, lawmakers and former officials raised concerns about potential damage to services ranging from extreme weather responses to efforts to prevent objects from colliding in space.” “The Hill: Energy – Environment – NOAA cuts”
What should be considered when making such a decision is what other services were prevented by directing these resources to NOAA activities rather than alternative uses. Even if the government just increases it overall budget, the added taxes or borrowing will have alternative uses.
You will immediately understand the issue when you consider your own household budget. Your income is limited (unless you give up some leisure to work more hours). You might gain some pleasure spending more on X, but you can only do so by giving up some Y. If you benefit more from the extra X than you lose giving up Y, then you should do it. It passes the cost/benefit test of maximizing the benefit of your given income.
In short, the fact that cutting the budget of some agency will cut some of its services is an incomplete argument for not cutting because if fails to take account of the rest of the cost/benefit assessment of the resulting reallocation of resources.
Such budget decisions are generally debated in Congress as it approves the government’s budget. It’s an imperfect process, like most of life, but it allows all views and pros and cons to be heard and considered. A body like Musk’s DOGE might be appropriate for evaluating the efficiency with which services are performed (perhaps proposing better information processing systems) and detecting corruption, but not for evaluating the desirability of such services themselves.