ANALYSIS OF EXPECTED PROGRAM CLAIMS AND ADMINISTRATION EXPERIENCE
Submitted to
Submitted by
Project Director: Ting Zhang
Lead Researchers: Lin Xiu; Dong Chen; Claire Guo; Adebamarajo Olateru-Olagbegi
Project Team Research Staff Members: John Janak, Sang Truong
This report includes an actuarial report in Appendix III from Milliman, Inc.
January 30, 2024
MARYLAND FAMILY AND MEDICAL LEAVE INSURANCE (FAMLI) PROGRAM-- PHASE II
This report provides a comprehensive analysis of expected program claims and administration experience for the Maryland Family and Medical Leave Insurance (FAMLI) program. It studies and projects the expected volumes, costs, and durations of claims by leave types for 2026-2030, and expected employer opt-out and the opt-in of self-employed individuals. Read the Executive Summary.
In the second stage of Phase II, we are tasked to conduct an analysis of expected
program claims and administration experience by studying and providing projections
on:
Read Chapter 1
Chapter 2 - Simulation of Claims and Costs using the USDOL Modified Worker PLUS Model
2.1 Simulation Methodology
The expected program claims, costs, and durations are simulated based on the modified
Worker Paid Leave Usage Simulation model developed by the U.S. Department of Labor
(USDOL Worker PLUS). We modified the USDOL Worker PLUS model to comply with the provisions
of the Maryland Family and Medical Leave Insurance Program (FAMLI). We used the DOL
Family and Medical Leave Act (FMLA) Employee Survey public microdata to train models
for individual workers’ leave needs and then draw individual Maryland workers’ characteristics
from 2017-2021 five-year American Community Survey (ACS) Public Use Microdata Sample
(PUMS) to simulate individual Maryland workers’ leave-taking behavior. The modified
USDOL Worker PLUS simulation model thus “runs” each sample person from Maryland ACS
data to predict his/her probability of taking a leave and the leave length based on
the Maryland FAMLI program parameters including eligibility rules and maximum leave
length, as well as FMLA-based distributions. The benefit costs are further simulated
based on the wage replacement structure, predicted leave length, and individual characteristics
from the ACS microdata.
Read Chapter 2.