MBAX 6007 Week 11 Final System Dynamics Paper Example

Hiring That Ignores Its Own Pipeline: A System Dynamics Explanation of Hearthside Home Care's Yearly Caregiver Swing, a Recommended Rule and Three Predictions the Agency Can Check

Student Name

Walden University

MBAX 6007: Complexity and Systems Thinking Dynamics

Instructor Name

Month Day, Year

Hiring That Ignores Its Own Pipeline: A System Dynamics Explanation of Hearthside Home Care's Yearly Caregiver Swing, a Recommended Rule and Three Predictions the Agency Can Check

The Pattern

For three years, Hearthside Home Care, a composite western Michigan business that sends home-health aides to older clients, has watched its roster of active aides rise above the number it needs and fall below it, roughly once a year. In the spring it refuses hospital referrals for lack of aides; by late summer it has more aides than visits and cuts everyone's hours; by winter enough aides have left that it is short again. Each phase has had its own explanation, weak recruiting in one season and poor scheduling in the other. This paper argues that both are the same structure seen at different points, and that the structure can be changed.

The Dynamic Hypothesis

The swing is produced by a hiring rule that responds to the visible gap between need and last month's roster while ignoring aides already in screening and training, combined with three delays: about a month for managers to perceive a shortage, about nine weeks for a new hire to become active and several weeks for aides to resign after their hours fall. Every later section supports or tests this claim.

The Model

The model has two stocks of aides, a pipeline and an active roster, and a third stock of referral memory among discharge planners. Hiring fills the pipeline; completions move people into the roster at a rate set by the pipeline's average length; quits drain the roster, starting from 4 percent monthly, rising in surpluses as short hours push aides toward steadier work and in shortages as overtime builds fatigue (Sterman, 2000). Refused referrals build referral memory, which lowers future referrals and fades over about six months. Four feedback loops follow: a balancing hiring loop, a reinforcing overtime and fatigue loop, a balancing short-hours loop that tends to overshoot and a referral loop that turns a shortage into a later drop in demand. The boundary keeps reimbursement rates, competing wages and population outside as inputs; with all of them held flat, the swing persists, which is the evidence that it is generated within the agency.

Confidence in the Model

Three tests support the model's use for policy. With hiring stopped, the roster drains steadily at the quit rate with no oscillation, as it should. With the outside world held constant, the base run still overshoots, peaking near 136 aides against a need of 120 and falling to about 110 eight months later. And the run's shape and timing resemble the agency's history, though the real swing has damped less than the model's, a gap partly closed when short-hours and fatigue quits were added. Forrester and Senge (1980) argued that confidence in a model comes from passing many such tests rather than any single one, and that it is never complete.

Policy Tests

Four remedies were tested under identical conditions. Results are illustrative but the ranking held across runs.

Bonuses illustrate what Senge (1990) called fixes that fail: they fill an already full pipeline and enlarge the next surplus. The overtime cap reveals a shifting-the-burden pattern, since overtime has been covering shortages while eroding retention, and removing it abruptly leaves visits unserved. The hours floor keeps aides through surpluses but slows the drain, so the next shortage arrives with an empty pipeline. Counting the pipeline addresses the cause directly.

Table 1

How Each Remedy Changed the Base Run Over 36 Months

Invented agency; extended-model runs with illustrative values.

Why Managers Miss the Pipeline

The failure to count the pipeline is not a sign of careless managers. In experiments run by Sterman (1989), participants managing a simulated supply chain kept ordering as though shipments already on the way did not exist, even when that information was in front of them, and produced swings very like Hearthside's. The aides in screening and training are equally easy to overlook: they appear on no visit schedule, draw no complaints and show up in no monthly report of unfilled visits. Making them visible on the report is therefore as important as changing the formula.

Recommendation

Hearthside should change its monthly hiring calculation. Each month, managers should estimate the number of aides needed, subtract the active roster and also subtract the aides currently in screening and training, discounted for the share who typically drop out, before adding expected quits. The result is the number to hire. The change requires one additional line in the monthly staffing report and no new spending. Overtime should then be reduced gradually as the swing narrows, not capped at once, and the agency should tell its two largest hospital partners about the change so that referral memory begins to fade.

Three Predictions

If the hypothesis is correct and the rule is adopted from the start of the next quarter, three things should follow. First, the next shortage should be shallower: the lowest monthly roster should fall less than half as far below need as last year's low point did. Second, it should come later, since the overshoot that sets it up will be smaller. Third, overtime hours should fall by at least a quarter within two quarters, because fewer and smaller shortages need covering. If none of these appears within a year, the hypothesis should be reconsidered, beginning with the referral loop, the least tested part of the model.

Limits

The model's parameters are illustrative, its fatigue and referral memory stocks are estimated rather than measured, and seasonal swings in hospital discharges are held flat. A real implementation would need Hearthside's own data on pipeline length, dropout and quits. The recommendation does not depend on any one parameter, however: in every run tried, counting the pipeline reduced the swing (Sterman, 1989).

Conclusion

Hearthside does not have a recruiting problem in the spring and a scheduling problem in the autumn. It has one problem, a hiring rule that cannot see its own pipeline, and that problem has a cheap fix the agency can test within a year.

References

Forrester, J. W., & Senge, P. M. (1980). Tests for building confidence in system dynamics models. TIMS Studies in the Management Sciences, 14, 209-228.

Senge, P. M. (1990). The fifth discipline: The art and practice of the learning organization. Doubleday.

Sterman, J. D. (1989). Modeling managerial behavior: Misperceptions of feedback in a dynamic decision making experiment. Management Science, 35(3), 321-339. https://doi.org/10.1287/mnsc.35.3.321

Sterman, J. D. (2000). Business dynamics: Systems thinking and modeling for a complex world. Irwin/McGraw-Hill.