מסמך טכני מיומן המחקר · באנגלית, כפי שנכתב בשלב העבודה
Step 296: partial demand-gap support
Vary the fraction of the previous year's underlying demand gap used to set support before the existing spending cap. Compare fractions 0, 0.25, 0.50, 0.75 and 1.00 at four investment-decay anchors and two rate sensitivities, with 432 matched cases per cell (17,280 simulations). The shared 15% positive-profit levy ceiling, benefit priority and the spending cap of 5% of output are held fixed. A joint pass requires the existing 90% consumption floor for both displaced groups and peak net public debt at or below 90% of output.
| Investment decay | Rate sensitivity | 0 | 0.25 | 0.50 | 0.75 | 1.00 |
|---|---|---|---|---|---|---|
| 0.000 | 0.60 | 297 | 297 | 297 | 296 | 289 |
| 0.000 | 0.15 | 265 | 268 | 268 | 267 | 262 |
| 0.518 | 0.60 | 269 | 277 | 280 | 285 | 281 |
| 0.518 | 0.15 | 261 | 282 | 285 | 291 | 283 |
| 1.270 | 0.60 | 60 | 110 | 143 | 144 | 134 |
| 1.270 | 0.15 | 27 | 75 | 88 | 90 | 91 |
| 1.500 | 0.60 | 22 | 50 | 66 | 64 | 60 |
| 1.500 | 0.15 | 8 | 35 | 38 | 38 | 41 |
Both 0.25 and 0.50 retain every joint pass from no support across these eight settings. At decay 0.518 and sensitivity 0.60, 0.50 recovers all seven cases lost under full support and gains 11 relative to no support. Its 280 passes are one fewer than full support. At 0.75 there are 19 gains and three losses relative to no support, with four of the seven full-support losses recovered. At sensitivity 0.15 at the same decay, 0.75 gains 30 without losses, retaining all 11 full-support losses as passes. At decay 1.5 and sensitivity 0.60, 0.50 has the largest joint count on this grid. At decay 1.5 and sensitivity 0.15, full support has the largest count.
These are comparisons among discrete policy fractions under the v293 assumptions. The absence of joint losses at 0.50 applies to the tested cases and threshold definition. It does not imply consumption or debt improves individually in every case or that this fraction is optimal beyond the grid.
Run python scripts/israel_investment_intensity_v296.py to reproduce counts, matched transitions and recovery of full-support losses. The endpoints reproduce the v293 joint counts and v295 full-support transitions.