The Stream by River Logic

River Logic Can Help Congress Balance the Budget

February 12, 2019 | By Eric Kelso

If only Congress would use River Logic

Watching the recent budget negotiations on TV, it is clear to me that Congress is currently grappling with two critically important issues: (1) infeasibilities – and how to avoid them; and (2) opportunity values – and how to get them and use them effectively. These same elements happen to be fundamental analyses available within River Logic's optimization modeling technology.

Our Congressional legislators – public perception notwithstanding – are not ignorant fools. They’ve known for a long time that a budget is infeasible when outflows (spending) exceed inflows (revenue, in the form of taxes, tariffs, selling government bonds, etc.)

Yet until very recently, Congress and the Executive branch, has acted as if they didn’t care enough to fix this problem, even though long-term deficit spending has been universally considered unwise and even reckless. Fixing the budget seems a proposition to which most of us can agree.

Although proposals and suggestions are now being discussed, it is too early to know how Congress will end the budget crises and lower the deficit, if not eliminate it altogether. Certainly, there will be some compromise – a feasible solution at least – that will eventually get passed.

But is there an optimal solution?

Yes…at least in the context of River Logic's capabilities. Congress could use River Logic to help balance the budget and dramatically reduce the national debt. Essentially, River Logic would allow them to turn the budget infeasibility flag back to “off” so they could solve for an optimal fiscal solution.

Since we live in a democracy, and I’m a voting, tax-paying citizen, I’m hereby offering my $0.02 for how to use River Logic to solve the nation’s budgetary crises. I’m certainly no political authority; actually, the opposite is true; however, I’d like to put forth an “out-of-the-box” solution for our nation’s top legislators. I fully admit this is quite a simplistic and optimistic illustration. Notably, I’m minimizing the significance of certain revenue-impacting decisions possible in the coming months, like simplifying the tax code by removing certain key deductions and credits. I’m also treating Congress as one decision-making body, when this exercise would actually necessitate separate scenarios run by both the House and the Senate.

Maximize Net Present Value

The model I propose will optimize on Net Present Value (NPV) for “Free Cash Flow.” To be considered a viable budget, each scenario’s total NPV should be positive. Yearly inflows will generally match all outflows (especially in later years); the goal being a debt-free, balanced budget over the long-term. The discount rate could vary by scenario depending on assumption (e.g., T-bill rate, inflation rate, GDP growth rate, etc.) The Horizon Value Method would be set to “Perpetuity.”


Time Horizon – Since there is no realistic possibility of reducing the deficit in the short term, the model must span a long-term time horizon, with many yearly time periods. Rather than focus on a single time horizon, it’s better to run various scenarios under various conditions with different time horizons (e.g., 10, 20, 30, 40 years, all of which I’ve heard discussed lately).

Costs and Limits – For a program to be funded (selected) it must be “purchased” for a cost. In fairness to all political parties, every program currently funded or under consideration will be represented at some level. If necessary, smaller, related programs could be combined into larger programs.

If just the consideration of certain programs ensures a healthy debate, the inclusion of minimum and maximum funding levels will certainly provoke serious arguments. An optimal solution, by definition, must honor all hard minimum and maximum constraints. So any program the government must purchase (i.e., minimum funding level) must be justified, most likely due to promised entitlements, required debt payments, or mandatory programs already funded in multi-year blocks that don’t require annual spending approval by Congress. And, as we’re hearing, even “required” funding levels are negotiable, thus throwing a wrinkle into what constitutes a constraint (more on this later).

As for the maximum funding level for each program considered, given the country’s current financial state it seems unlikely most numbers will exceed the 2011 budget numbers save for a cost-of-living or inflationary increase in future years. They could certainly be less than current levels. No program will get a blank check, no matter how popular it is.

Revenues – Inflows will simply be represented by the total revenue expected to be collected for each year. Various scenarios can reflect different revenue assumptions (e.g., “trickle down” economics, tax changes, etc.). Whatever the expected level of revenue for each year for that scenario, it will act to constrain the overall annual budget. 

Perceived Value – Program costs and expected revenues alone are not enough to drive meaningful solutions. If so, we would have a balanced budget today minus the bickering and name-calling. Above and beyond minimum funding levels, some measure of perceived value must drive any additional “sale” of each program.

This suggestion might seem impossible to implement, except that it exists today – it’s called politics. Every legislator inherently knows, or can discover, a program’s value to his or her constituents. As part of the data collection process, each would be required to assign a value for each program. This value weighting would rank each program from zero (no value) to 100 (must fully fund in budget). To ensure a somewhat equal distribution of values, each legislator must attempt to equally distribute all considered programs, with 25% of all programs falling into each value weighting quartile (e.g., 0-25, 26-50, 51-75, and 76-100). This would prevent “cheating” by placing all programs in either the first or last quartile, depending on one’s political ideology.

For each program, the sum of all member values becomes the total value in the model. With 535 total members in Congress, the minimum value of each program is zero, the maximum is 53,500.

Important Constraints

In case you are thinking, “That’s too easy. Just rank the programs from highest to lowest perceived value and cut them off when the revenue is gone!” The model can utilize these additional constraints when relevant for each program. Together, these integer constraints make it exponentially harder to solve without optimizing:

  1. Fund or don’t fund in any single year;
  2. If minimum funding level is required, let the solution choose to fund the program at that level or higher, or not at all;
  3. If funded in any year, then fund that program each subsequent year for the remainder of the time horizon;
  4. If defunded in any year, then forever sunset the program; meaning, it must stay defunded for the remaining time horizon.


Each scenario run will determine a list of funded programs and at what level. The collection of all scenarios will produce a list that can then be ranked by NPV. Scenarios with the lowest NPVs will likely be discarded, while those with highest NPVs will be used as a blueprint for the official budget to be voted on, approved, and presented to the President.

Importantly, unlike the last 235 or so years, opportunity values will be available to guide future budgets. Those programs purchased at the minimum level will likely have negative opportunity values – meaning, there’s likely a better use for that money in the future. Conversely, programs purchased at the maximum level will likely have a positive opportunity value – meaning, they should be considered for additional funding if future funds are available and the money can be well spent.

This approach will certainly not solve all our federal budget problems, but it’s a solid attempt. If anyone from Congress happens to read this blog, please give me a call. I’m ready to get started. Readers are also free to share this content with their congressmen/women. Let them know there is a solution available today which can positively impact the budgeting process!

Optimization or Similation



    Posts by Topic

      Supply Chain Brief