paper/diagnose.py takes inflow, capacity, backlog and target time per
link and returns utilization, throughput time and backlog growth per
link, the binding link, the type of measure in Theory of Constraints
order (exploit, subordinate, elevate) and, for the binding link, the
capacity needed under the coordinated rule from sim.py, sized on the
forecast flow rather than the queue.
The example file holds the stylized month-24 values from the coordinated
run and reproduces section 9.5 (utilization 1.28, +19 % capacity);
checked against simulate_endo("koord"), which orders +19 % in month 24.
sim.py and paper/data/ are unchanged. Time paths and the stability check
for the police rule are not in this version.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The slides now present the model as a tool right after the idea: a new
slide on what it finds (binding link and type of measure, how strong,
effect on the rest of the chain and on trust), the seven steps, and the
worked full-court example with the table from section 9.5 next to the
coordination figure. The findings follow as what the tool builds on. The
separate full-link slide is replaced by the example; 16 slides.
The font path for Inter is no longer a hard-coded Linux path, so the
slides build with TeX Live's copy of Inter.
main (57 pages), notat (2) and karusell (16) are rebuilt from the source
in this repository and replace the PDFs in pdf/. README, CHANGELOG and
CLAUDE.md state the new page counts and the verified build commands.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The abstract, introduction, conclusion and summary note now present the
model as a tool that goes from observed numbers to the binding link, the
type of measure (Theory of Constraints), how strong it can be (stability
condition) and its effect on the rest of the chain and on trust. The three
contributions are reframed as the three parts of that tool, and the
novelty table gets a first row for it.
New subsection 9.5 runs the tool step by step (a-g) on a full court and
compares its recommendation with the obvious alternatives (new table,
numbers taken from the existing runs). The summary note stays at two
pages. sim.py and paper/data/ are unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
New appendix C "Fra ligning til kode": for each of the 39 numbered
equations, the function and lines in sim.py that compute it, the figure
or table it feeds, and its status (18 implemented, 11 simplified, 10 not
implemented), plus a table tracing every figure and table to code and
data file. Documents small code-text differences (cap in the xi rule,
latched binding and dead zone in the coordinated rule, two-sided quality
loop). Introduction points to the appendix. sim.py and paper/data/ are
unchanged; main.tex builds with no undefined references.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Abstract: states first what is known and already done (dynamic optimal
enforcement, chain forecasting in NL and SE, Norway's planned forecasting
model and police priority analysis), then the three contributions with
the coordination result first: forecasts are not enough.
- Introduction: prior and ongoing work now precedes "Bidrag" and ends in a
table (tab:nytt) separating known in research / in use / under way in
Norway / new here. "Bidrag" is rewritten to the three new things; the rest
of the framework is labelled as mainly built on known work.
- Conclusion: one paragraph on what the result means for the chain
forecasting model the ministry expects, limited to the court link that
the model actually simulates.
- Summary (notat): new section "Hva som finnes, og hva som er nytt"; key
finding retitled "prognoser er ikke nok"; tightened to stay at two pages
(margins 16 -> 14 mm, figure height 2.8 -> 2.5 cm, shorter source line).
Built with pdflatex (TeX Live 2026): main 52 pages, notat 2 pages, no
undefined references. pdf/ is not replaced yet; that is the next part.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Forsøksloven (1992, amended 2021) already allows four-year trials in public
administration, but only for organisation and task allocation and never by
restricting rights; trials with procedure and sanction forms need their own
legal basis.
- A general sunset act for regulations was rejected by the Storting in 2001
(Innst. O. nr. 77 (2000-2001)); the proposal here covers intrusive powers only.
- Riksrevisjonen, Dokument 3:13 (2007-2008), found deadline breaches across
several links of the chain.
- NOU 2020:11, p. 197, records a minority of the previous court commission
warning that one oversized link creates bottlenecks.
All four were read in the primary source (Lovdata, stortinget.no,
riksrevisjonen.no, regjeringen.no PDF). Not built: TeX not installed yet.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
One sentence in "Hva modellen er" so the two-page summary does not claim
more than the main text: the Netherlands and Sweden already run whole-chain
forecasting models, and the ministry expects one for Norway. Sources added
to the source line. Not built: no TeX on this machine; the two-page limit
must be checked in the PDF build.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The prior-work paragraph claimed novelty for the coupling only, but left out
three bodies of work a reviewer would raise:
- dynamic optimal law enforcement (Feichtinger and Tragler 2002; Tragler,
Caulkins and Feichtinger 2001; Grass et al. 2008)
- whole-chain forecasting models already used for justice budgets
(Netherlands PMJ since 2004, Swedish joint forecasts since 2008,
England and Wales MoJ projections, NSW simulation model)
- Norwegian work under way (joint chain statistics until a forecasting
model is ready, 2023; police priority-effectiveness assignment due
31.12.2026; Riksadvokaten's priorities, stable over time by its own account)
Also adds harm-focused policing (ONS 2016, Sherman 2013, Ratcliffe 2015),
risk-based regulation (Black and Baldwin 2010), Pontell 1985, Olaya et al.
2018, Hao et al. 2026 (preprint, marked as such) and Azaria et al. 2023. The
contribution is restated as three things: the stability analysis of the
delayed loop, the coupling in one model, and the shared-constraint method.
Every new reference was checked against Crossref or the publisher's or
issuer's own document. pdf/ is not rebuilt: no TeX installation here.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The note said most estimates lie between -0.1 and -0.5 (citing Chalfin and
McCrary 2017). The review instead gives a best guess of about -0.4 (violent)
and -0.2 (property), with quasi-experimental estimates spanning roughly -0.1
to -2, few of them significant (JEL 55(1), pp. 14 and 37). The conclusion
that the model's -0.83 is at the strong end still holds.
pdf/ is not rebuilt: no TeX installation on this machine.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
PDFs and LaTeX source are CC BY 4.0; sim.py, requirements.txt and the
figure data stay MIT. README and CLAUDE.md state the split.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Concept note "Den dynamiske rettsstaten" version 2: PDFs, LaTeX source,
simulation (sim.py) and figure data. Working documents are kept out of
the repository.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
