Allocating Powers Between National and Local Government: Reserved, Enumerated and Shared Competences

October 2026

Download: PDF · Word (DOCX)

Series: Local Government, Civil Service & Public Administration — research papers accompanying the the Local Government working blueprint. Paper LG-01.

1. The question

When the national government acts in a field where a locality already acts, which level’s decision governs — and who decides where the boundary lies?

Without an explicit allocation rule, the default is drift: the national level accumulates competences through emergencies, fiscal leverage, and administrative precedent, while local governments become implementers of decisions they did not make. The PRC (Mainland) illustrates the unallocated extreme — its constitution provides no division of power among tiers of government (source). This paper proposes the opposite for a free China: a written, public allocation of reserved, enumerated, and shared competences.

2. The blueprint’s current position

This paper takes its position from the blueprint page itself, Local Government, Civil Service & Public Administration. In its “Reserved and enumerated powers” section, the blueprint states its working direction: “to protect ordinary local governance while enumerating functions that truly require national treatment.” It adds three working rules. First, cross-regional cooperation must distinguish information exchange, jointly operated services and binding delegated joint authority — coordination does not automatically create command authority. Second, each joint mandate should identify its members, scope, voting or decision rules, finance, duration, withdrawal and dispute resolution. Third, “fiscal power should follow lawful responsibility rather than become a hidden mechanism for centralizing functions.” The blueprint carries a standing caveat: it states a current design direction, not a final constitutional settlement, and exact allocations remain subject to constitutional and legal research and future lawful representative decision. This paper develops that direction into a concrete allocation design.

3. Evidence and comparative options

3.1 Germany — exclusive and concurrent legislation, residual power with the Länder

Germany’s Basic Law divides legislative competence into three tiers. Article 73 lists the Federation’s exclusive powers — foreign affairs and defence, citizenship, currency and money, the unity of the customs and trading area (source). Article 74 lists concurrent powers — civil and criminal law, road traffic, business law, consumer protection — where the Länder may legislate so long as the Federation has not done so (source). Beneath both sits the residual rule: Articles 30 and 70(1) assign state power to the Länder as the general rule, with the federal level legislating only where the constitution expressly grants it; the Länder also implement most federal laws (source).

3.2 The United States — enumerated federal powers, reserved remainder

Article I, Section 8 lists Congress’s enumerated powers — taxation, interstate and foreign commerce, coining money, declaring war, raising armies — limiting what the national government can do (source). The Tenth Amendment supplies the residual clause: “The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people” (source)). The cautionary lesson: without an effective residual rule, an enumerated list leaks — the Necessary and Proper Clause and two centuries of broad Commerce Clause interpretation steadily expanded the centre’s reach (source).

3.3 Switzerland — cantonal sovereignty as the default

Article 3 of the 1999 Federal Constitution guarantees cantonal sovereignty in every sphere where the Constitution does not explicitly provide for federal competence, and Article 42 restricts the Confederation to tasks the Constitution attributes to it — so every new federal competence requires a constitutional amendment approved by a majority of voters and a majority of cantons. The Federal Supreme Court has confirmed the cantons’ residual general competence: everything not conferred on the Confederation stays cantonal (source). Cantons implement federal law, conclude agreements among themselves, and participate in federal decision-making through the Council of States and mandatory referenda (source). Switzerland is the purest “coming together” federation: enumerate the centre, and the units keep the rest (source).

3.4 ROC (Taiwan) — self-government matters versus commissioned matters

The closest operating precedent for a Chinese polity is ROC (Taiwan). The Constitution’s Chapter 11 provides for county and special-municipality self-government, and the Local Government Act distinguishes two kinds of local business: self-government matters, carried out in the locality’s own right, and matters commissioned by higher government agencies, implemented on the centre’s behalf (source). The Act enumerates self-government matters in detail — local elections and recall, land registration, local finance and taxation, social welfare, education, urban planning, environmental protection and health administration (source). The distinction carries a supervision rule confirmed by the Constitutional Court: for self-government matters central supervision is confined to legality review, while for commissioned matters the centre may also supervise appropriateness (source). Authority disputes among the central government, special municipalities, and counties/cities are resolved by the Legislative Yuan (source). This offers three usable ideas: an enumerated self-government list, a formal own-business/commissioned-business distinction, and tiered supervision.

3.5 The subsidiarity principle — the burden of justification points downward

Article 5(3) of the Treaty on European Union defines subsidiarity: the Union acts only where objectives cannot be sufficiently achieved by the Member States at central, regional, or local level but can better be achieved at Union level by reason of scale or effects; decisions are taken as closely as possible to the citizen (source). The related conferral principle keeps policy areas not explicitly agreed in the Member States’ domain (source). Subsidiarity entered European primary law at Germany’s insistence precisely as a barrier to the exercise of competences — a check on centralisation. The usable core: when the national level proposes to act in a shared field, it must show why the local level cannot achieve the objective — the burden of justification points downward.

3.6 Options on the table

Three tested architectures emerge: (i) the enumeration-centre model (Germany/US) — list national powers, reserve the rest to localities, add a concurrent tier where enacted national law prevails; (ii) the sovereignty-default model (Switzerland) — new national powers need a double-majority amendment, the strongest firewall against competence creep; (iii) the two-list model (ROC (Taiwan)) — enumerate both self-government and commissioned matters, with legality-only supervision for the former.

4. Recommended design: numbered provisions

P1. Constitutional enumeration of national competences. List the national government’s exclusive competences: national defence, foreign affairs, citizenship and immigration, currency and the central bank, customs and the unified national market, and inter-regional criminal-police cooperation. No national act may claim a competence outside the enumerated list.

P2. Residual competence belongs to local government. All public competences not enumerated nationally and not assigned to the shared tier in P3 are reserved to the provincial (or equivalent) level, which may devolve them downward by local statute. The residual is a constitutional right of local government, not a delegation the centre can withdraw by ordinary law.

P3. Shared (concurrent) tier with a written list and a precedence rule. Name the shared fields in the constitution — civil and commercial law, criminal law, labour and social insurance, environmental protection, transport networks. In shared fields, localities may legislate unless national legislation occupies the field; where it does, national law prevails only to the extent of the actual conflict.

P4. Self-government matters versus commissioned matters. Following the ROC (Taiwan) distinction, classify every local function as a self-government matter (local budget, local taxation within national bands, local elections, land use and urban planning, basic education, local policing, local culture and welfare) or a commissioned matter (national policies the locality implements as the national government’s agent, funded by the national government). National supervision of self-government matters is confined to legality review; supervision of commissioned matters may extend to appropriateness — but commissioned tasks must be funded. No unfunded mandates.

P5. Subsidiarity procedure. Any national proposal touching a shared or reserved field must carry a written subsidiarity statement explaining why the objective cannot be sufficiently achieved locally. The statement is reviewable by the constitutional court at the suit of an affected local government.

P6. Competence disputes go to a court, not to the executive. Disputes over tier assignment are decided by the constitutional court (or, during transition, the review body designated under the blueprint framework).

P7. Competence-creep lock. Adding a new exclusive national competence requires a constitutional amendment; expanding national legislation within the shared tier requires a statute that survives the P5 subsidiarity statement. Framework legislation in the shared tier shall prefer minimum standards over exhaustive regulation, preserving room for local variation.

P8. Fiscal correspondence. Competence without revenue is decoration. Each assignment in P1–P4 must be accompanied by a revenue assignment sufficient to fund it, plus a fiscal equalisation mechanism for poorer localities.

5. Open questions and risks

Q1. Provincial tier or county-direct? Whether the primary local tier should be provinces or large municipalities is not settled here — it belongs to LG-12 (“Who Decides Who Governs What?”), which this paper does not pre-empt.

Q2. How to prevent the American leak? The United States shows enumerated lists expand through broad interpretation (source). The P7 lock is the proposed answer, but it depends on a constitutional court willing to strike down national overreach — an institution that must be built and staffed, per the activation-ledger principle of the companion constitutional-transition blueprint (source).

Q3. Transition sequencing. A transition government will necessarily exercise broader emergency powers; the risk is that emergency centralisation hardens into permanent allocation. Mitigation: every emergency competence grant carries a sunset date and is reviewed under the P5/P6 procedure at expiry — no grant may be renewed on the basis of its own precedent.

Q4. Fiscal capacity gap. Poorer localities may lack the revenue base for reserved competences, creating pressure to re-centralise. P8 names the problem; the equalisation formula is an open design task for the fiscal papers.

Q5. Minority and border regions. Competence allocation interacts with autonomy arrangements for regions with distinct ethnic or historical identities — a question requiring separate treatment under the relevant autonomy provisions.

RISK — Paper before institutions. An enumerated schedule means nothing if no court enforces it and no local legislature exists to use it; the allocation must be tracked from text through institutional seating to first accountability interaction (source).

6. Cross-links

  • The Local Government blueprint — Local Government, Civil Service & Public Administration: this paper develops the blueprint’s “Reserved and enumerated powers” direction (§2) into the allocation design of §4.
  • LG-12, “Who Decides Who Governs What?” — the companion paper on which tier holds primary local competence; this paper’s P2 deliberately leaves that choice to LG-12.
  • CT04 federalism history 1898–1947 — the record of Chinese federalist experiments (provincial constitutions, the 1920s federalist movement) is referenced for context on why competence allocation failed in the early Republic; this paper does not rewrite that history.
  • Sibling LG papers on fiscal equalisation, civil service, and local elections are expected to take up P8, local staffing, and local electoral legitimacy.

Also published in this series: County, Municipal and Provincial Transition Bodies · Essential Services Never Stop · A Professional Civil Service, Not a Faction’s Staff · Administrative Accountability in Practice.

Working paper v1.0, published October 2026 as part of the Local Government, Civil Service & Public Administration blueprint program. This paper will be updated as the evidence base develops.