Week 8 Homework Feedback: Pablo Romanella

Assignment: Dynamic Programming & Value Function Iteration with Population Growth
Week: 8
Date: Week 8 Assessment

⚠️ Overall Assessment

Result: ⚠️ Partial <50% Correct

The code base is neat and modular, but it solves the (n=0) model from Week 7. The helper solveVFI constructs cons = f(k_i) + (1-\delta)k_i - k'_j and the simulation uses (c_t = f(k_t) + (1-\delta)k_t - k_{t+1}), so the ((1+n)) population-growth term never enters. Consequently, every figure and interpretation corresponds to the wrong homework. 5/15 tasks correct (≈33%).

Main Problems

  • Task 1 (consumption matrix) fails: the ((1+n)) multiplier is missing everywhere; (n) never appears after the parameter block.
  • Simulation (tasks 6–9) uses the wrong resource constraint: consumption ignores the extra investment needed to equip new workers.
  • Parameter experiments only vary (\alpha) and (\beta) under the (n=0) specification, so they do not speak to the Week 8 question.

What to Fix

  1. Pass (n) into solveVFI and build 
    cons = fvec + (1-delta).*kgrid - (1+n)*kgrid';
  2. Simulation: use c = f(k) + (1-delta)k - (1+n)k' and recompute all transition plots.
  3. Comparative statics: add at least one (n) experiment (e.g., (n = 0, 0.01, 0.02)) to illustrate the dilution effect discussed in the slides.

Once those changes are in place, re-run the notebook so we can reassess against the Week 8 rubric.