Problem solving a Ramsey Problem with 2 instruments and 1 objective function (sum of agent's utilities)

Hi all,

I am solving a Ramsey problem in a model of 3 countries.
The model works fine when based on simple policy rules.

Now, I want to make 2 of the countries solve a Ramsey problem where the objective function is the sum (or average) of their utilities. However, I am running to an error:

If I comment one of the rules and use the Ramsey problem solver as follows it works:

// Objective function and discount factor
planner_objective Util1;
ramsey_model(instruments=(taui), planner_discount=betta);

However, when I comment the other rule (tool) and try to obtain a cooperative Ramsey solution with:

// Objective function and discount factor
planner_objective Util1 + Util2;
ramsey_model(instruments=(taui,taue), planner_discount=betta);

I get that the Blanchard Kahn conditions won’t hold:

There are 73 eigenvalue(s) larger than 1 in modulus
for 72 forward-looking variable(s)

The rank condition ISN'T verified!

Error using print_info (line 42)

Blanchard Kahn conditions are not satisfied: no stable equilibrium

has someone any clue what is the problem?. I checked the timing of the variables and even with a toolbox I am able to get the nash solution but the joint welfare maximization is the one causing me trouble (both using dynare++ or with the toolbox by Bodesntein et al (2019,JME)).

PS. As an experiment I am running it with only one Utility function in the objective function:

planner_objective Util1;
ramsey_model(instruments=(taui,taue), planner_discount=betta);

this case (not exactly the one I am interested but one using two tools as well) works fine.

Something else, it looks that the problem is one of the lagrange multipliers created by the Ramsey problem solver(the mismatch between eigen values and non-predetermined variables is only 1), i.e., apparently, my problem relies in only 1 variable that is causing trouble. There may be a redundant forward looking variable that leads to the mismatch between the #of_eigen>1 and the #of_forward looking variables.
MaP3iec_Notworking.mod (7.8 KB)
MaP3iec_working.mod (7.8 KB)

My answer at


also applies here.