Step 293: fiscal sensitivity to the Israeli investment anchors

This reruns the v286 grid with two descriptive estimates from v291. The v282 mapping divides the negative slope by 0.34 and clips the result to [0, 1.5]. The full sample with 2012 break controls gives −0.5443, which maps to 1.601 before clipping and 1.500 in this exercise. The post-2012 estimate of −0.1761 maps to 0.518. The mapped confidence intervals span [0.629, 1.5] and [0, 1.5] respectively. These clipped ranges are descriptive transformations of the empirical intervals. The regressions cover nonhousing fixed investment, including government investment, whereas the model parameter affects autonomous planned investment. The mapping is illustrative and has no causal interpretation.

Every table entry counts joint passes out of 432 scenarios. Both displaced groups must maintain at least 90% of initial real nonhousing consumption in every year. Peak total net public debt must stay at or below 90% of output. The support policies use the v285 shared 15% positive-profit ceiling and priority for the benefit levy. Entries report the two interest-sensitivity assumptions, 0.60 / 0.15.

At both Israeli anchors and the OECD benchmark, gap-triggered support improves the joint count at both interest sensitivities. At zero decay it reduces the count, so the ranking is conditional on the amount of investment weakness. At the full-history stress setting, even the best policy yields at most 60 joint passes. At 0.518 and interest sensitivity 0.60, gap support raises the joint count from 269 to 281 while standalone debt passes fall from 356 to 354. Thus the joint ranking cannot be inferred from the debt pass count alone. At 1.27, the run reproduces the v286 shared-levy results. Grid shares represent the selected scenarios and carry no probability interpretation.

Run python scripts/israel_investment_fiscal_v293.py for the mapping, joint counts and separate consumption and debt counts.