Policy and Complex Systems - Volume 2 Number 1 - Spring 2015

William G . Kennedy , ^A Emily S . Ihara , ^A Catherine J . Tompkins , ^A Megumi Inoue , ^A and Michael E . Wolf-Branigin ^A

Caregivers providing support to family members with Alzheimer ’ s disease often encounter high levels of stress within the fragmented long-term care system . To address this emerging issue affecting millions of families , we applied agent-based computational modeling methods to better understand the impacts of policy alternatives . Potential options include increased respite care , tax incentives , work place policies , and adult day services as alternatives to reduce caregiver stress . Experiments with our model demonstrate that policy options providing programs , services , and support for caregivers can reduce their stress by providing a minimum of 16 hours per week of respite care .

Keywords : agent-based modeling , caregiving , complex adaptive systems , respite , older adults , social policy

multidimensional issue that will continue to affect millions of families . A likely contributor to caregiver stress is the fragmented long-term care system that has developed over many years to address specific problems without an overarching directed effort to create a seamless wraparound approach . Policymakers cannot agree on the best policy solutions , particularly to fund long-term care services . Although there are no easy answers , agentbased modeling provides a computational method to forecast individual and group interactions occurring within dynamic systems , and therefore is an ideal method for testing various policy solutions for a complex issue . The increased computational power available to even the casual user over the past decade facilitates using this bottom-up approach for creating simulated environments in which experiments may be conducted . Agent-based modeling provides an approach for investigating complex phenomena by computationally simulating the interactions of autonomous agents in order to assess their effect on whole systems . Agent-based models ( ABMs ) create a “ social reality ” generated from several variables or inputs . ABMs are created to model an environment in which interactions , characteristics , and behaviors of individual agents are identified ; this leads computational simulations to forecast the emergent behavior for the entire group .

A

George Mason University , Fairfax , Virginia 10.18278 / jpcs . 2.1.5

31