214 Policies, an Expiring Telework Rule, and a Retirement Cliff

Sequence by hard deadline, legal exposure, and dependency - not by which initiative is most exciting.

First (0–60 days): the telework policy. It has an external deadline, affects 38% of staff, and is almost certainly a mandatory subject of bargaining. An expiration with no replacement creates inconsistent practice that becomes past practice, and unmanaged expectations. Deliverable: a legally reviewed, bargained, time-limited policy with a scheduled review date.

Second (30–120 days): the 12 CBA conflicts and the 6 statutory conflicts. These carry live liability - a supervisor applying a conflicting policy can generate a grievance the agency will lose. Identify them by name, correct or suspend them, and notify affected managers.

In parallel (60–180 days): workforce plan data collection. Retirement eligibility and skills-gap data take months to assemble; start now so the plan is evidence-based rather than aspirational. The policy audit and the workforce plan share the same underlying data (classifications, supervisors, tenure) - build the extract once.

Deferred (180+ days): the routine five-year review backlog. It matters, but it is not the emergency.

Tell the leadership team this sequence explicitly. Two analysts cannot run three priorities at full speed, and pretending otherwise guarantees all three slip.

Do not launch a mass rewrite; it will consume both analysts for a year and produce policies nobody reads. Build a system.

Prioritize by risk tier

• Tier 1 - legal exposure: conflicts with statute or CBA, anything touching pay, leave, discipline, or safety.
• Tier 2 - operational: telework, attendance, technology use, scheduling.
• Tier 3 - administrative: travel, forms, housekeeping. These can be handled by batch review or retirement.

Govern with five controls

1. A policy on policies - who drafts, who reviews (legal, labor relations, budget), who approves, how employees are notified.
2. A named owner and a mandatory review date on the face of every policy. No undated policies.
3. Sunset by default - anything not reaffirmed within its cycle lapses. This solves the 41% problem permanently.
4. A single index with version control, so a supervisor cannot find three versions of the attendance policy.
5. Union notice built into the workflow, not added at the end.

Also expect to retire policies, not just revise them. A 214-policy manual in a 2,100-person agency almost certainly contains 40–60 documents that should be consolidated or eliminated outright, and retiring them is faster than rewriting them.

Plan content - keep it short enough to be used:

• Gap analysis by classification and location: retirement eligibility, attrition trend, and time-to-fill against projected demand.
• A critical-role list (typically 15–25 positions, not 200) where retirement risk and operational impact overlap.
• Knowledge-transfer and succession actions tied to those roles, with named successors or a stated 'build vs. buy' decision for each.
• Hiring and development pipeline commitments with dates and owners.
• A monitoring section: metrics reviewed quarterly with the budget cycle.

Metrics that will actually be tracked: percentage of critical roles with an identified successor, retirement-eligible percentage by unit, time-to-fill for critical classes, internal fill rate, and skills-gap closure.

Handling the pace expectation Give leadership early, visible wins that are real: publish the critical-role list at 60 days, retire 25 obsolete policies at 90 days, and issue the telework policy on time. Those are demonstrations that the system works, not theater. Be direct about capacity: with two analysts, the plan will be built in phases, and the fastest way to lose a year is to promise a comprehensive plan in one quarter. Frame the retirement cliff in cost terms - a supervisor vacancy that takes 228 days to fill is a budget event - because that argument survives leadership turnover and budget cycles in a way that 'good HR practice' does not.