Continuity of Operations for Local Government: A Plan You Can Execute

Most local governments plan carefully for emergencies affecting the community and thinly for emergencies affecting themselves. The EOP describes how the county responds to a flood. It rarely describes what happens when the flood takes out the building where the finance office, the permitting system, and the 911 administrative function all live.

Continuity of Operations planning fills that gap. It is less dramatic than response planning and, in terms of days of disruption avoided, usually higher value. The events most likely to interrupt a local government are not catastrophes. They are a building fire, a ransomware incident, a water main break under the courthouse, or a staffing shortfall.

The gap most jurisdictions have

Three patterns show up repeatedly:

• The plan exists but names everything as essential. If every function is essential, the plan gives no guidance when you can only run four of them.
• The plan assumes a facility problem only. It describes relocating to an alternate site, which does not help when the building is fine and the network is encrypted.
• The plan was never tested. Nobody has checked whether the alternate site has enough network drops, whether staff can actually access systems from it, or whether anyone knows where the vital records are.

Identifying essential functions honestly

An essential function is one that must continue, or resume quickly, because stopping it causes legal, financial, or safety consequences. Discipline here is the whole exercise.

Useful questions for each candidate function:

• What happens if this stops for 24 hours? For a week? For a month?
• Is there a statutory or contractual deadline attached to it?
• Does stopping it put anyone at physical risk?
• Does stopping it stop payments in or out?

Most counties land on somewhere between eight and fifteen genuinely essential functions. Typically: emergency communications and dispatch, law enforcement and fire response support, emergency management coordination, payroll, accounts payable, utility and public works operations, public health functions, court operations with statutory deadlines, vital records issuance, jail operations, and executive decision-making and public information.

Recovery time objectives and dependencies

For each essential function, record a recovery time objective: how quickly it must be running again. Then record what it depends on, because the dependency is usually the real constraint.

Two things fall out of this table that rarely surface otherwise. First, single points of failure in people: functions where exactly one person knows how to run the process. Second, dependencies you do not control, such as a vendor-hosted system whose recovery commitments may not match your RTO. Both are findings worth acting on regardless of whether a disruption ever occurs.

Succession and delegation of authority

Orders of succession answer who acts when the incumbent is unavailable. Delegations of authority answer what that person may decide without further approval. They are different, and plans frequently include the first and omit the second.

• Go at least three deep for every key position. Two deep fails in exactly the scenarios continuity planning exists for.
• Use positions, not names, with a roster appendix that is reviewed annually.
• State the limits. Spending authority, hiring authority, authority to close facilities, authority to speak for the jurisdiction. A successor who does not know their spending limit will not spend.
• Check the legal basis. Some delegations require ordinance, resolution, or charter authority. Have the county or city attorney confirm that what the plan claims is actually enforceable.
• Document how succession is triggered and communicated, so it is not ambiguous at the moment it matters.

Alternate facilities and remote work

Traditional COOP planning emphasizes a physical alternate facility. That still matters for functions that cannot be performed remotely, such as dispatch, detention, and public works. For office functions, distributed work is often the faster and cheaper continuity strategy, provided it has been tested.

For each alternate facility, verify rather than assume:

• Capacity, network connectivity, and power, including backup power.
• Whether the systems the relocating staff need are reachable from there.
• Access control: who holds keys or credentials, and at what hour.
• A written agreement if the facility belongs to someone else.
• Geographic separation sufficient that one event does not take both sites. An alternate EOC four blocks away does not help in a flood.

For remote work, confirm the specifics: licence counts for concurrent remote access, whether staff have devices at home, whether the phone system can route to them, and whether they can sign documents. These are cheap to check in advance and expensive to discover during an incident.

Vital records and systems

Vital records are what you cannot operate without: current financial records, personnel and payroll data, contracts and agreements, legal and property records, plans and system documentation, and the credentials needed to reach any of it.

For each, document where the authoritative copy lives, where the backup lives, how quickly it can be reached, and who is authorized to reach it. Then test a restore. An untested backup is a belief, not a control.

Activation, and the first four hours

The plan should be explicit about who can activate it, at what threshold, and what happens immediately. Partial activation matters: most real disruptions affect one department or one system, not the whole government.

A workable first-four-hours checklist covers: confirming personnel accountability, deciding activation scope, notifying successors if incumbents are unavailable, notifying staff of reporting instructions, establishing a communication channel that does not depend on the affected system, notifying key external partners and the public, and starting a cost and decision log from minute one.

That last item is routinely forgotten and routinely regretted. Documentation started on day three never reconstructs days one and two accurately.

Testing without pretending

Continuity plans fail in mundane ways that only testing reveals. Options, roughly in order of effort:

• Call-down test. Can you reach your successors and key staff outside business hours, using current numbers?
• Records retrieval test. Ask someone to retrieve a specific vital record from backup, timed.
• Remote access test. Have a department perform its essential function entirely from outside the building for one day.
• Alternate facility walkthrough. Go there, connect, and try to work. Take photographs of what is missing.
• Tabletop. Walk leadership through a scenario that removes a building or a system. Our tabletop guide covers design.
• Full relocation exercise. Move a department for a day. Expensive, revealing, and worth doing once every few years.

Findings go into an improvement plan with owners and dates, the same discipline described in our after-action report guide. A COOP plan that is reviewed annually and tested in some form every year stays true. One that is written and filed is accurate for about eighteen months.

Common questions

What is the difference between COOP and continuity of government?

Continuity of Operations focuses on maintaining an organization's essential functions. Continuity of Government is broader and concerns preserving the constitutional and statutory functions of government itself, including the ability of governing bodies to convene and act. Local jurisdictions usually address both, with continuity of government elements covering succession of elected officials and the ability to hold lawful meetings.

How often should a COOP plan be updated?

Review contact information, succession rosters, and system dependencies annually at minimum, and revise the plan after any significant organizational change, facility change, or major system replacement. Any disruption, exercise or real, should feed the next revision.

How long does it take to build a COOP plan?

For a mid-size county, expect roughly three to six months when departments participate actively. The business process analysis, meaning identifying essential functions, recovery time objectives, and dependencies with each department, is the bulk of the work. Writing the document is comparatively quick.

Do small jurisdictions need a COOP plan?

Arguably more than large ones, because small jurisdictions have fewer people who know how to run each process. A town where one clerk runs payroll, utility billing, and permitting has a more acute continuity problem than a county with depth in each function.