Aggregate Planning Costs and Trade-offs

Introduction

Aggregate production planning translates a medium-term demand forecast into a feasible and economically attractive production plan. The planning horizon commonly extends from three months to eighteen months. Decisions are made for product families rather than individual stock-keeping units, and the principal concern is the coordinated use of workforce, capacity, inventory and external supply.

Aggregate planning is not simply a cost-minimisation exercise. A plan with the lowest accounting cost may create unacceptable late deliveries, excessive employee turnover, quality risks, supplier dependence or strategic inflexibility. The central managerial problem is therefore to balance cost, service, capacity and workforce stability under uncertainty.

This chapter develops a cost model for regular-time production, overtime, hiring, layoffs, inventory, shortages, backorders, subcontracting, idle capacity and changeovers. It also explains the trade-offs among these alternatives and illustrates their use through an applied case.

Learning objectives

After studying this chapter, the reader should be able to:

1. Formulate an aggregate-plan cost model with consistent units and time periods.
2. Distinguish inventory, shortage, backorder, capacity and workforce costs.
3. Explain the economic trade-offs among level, chase and mixed strategies.
4. Construct and interpret a feasible aggregate production schedule.
5. Evaluate service, workforce and strategic consequences in addition to direct cost.
6. Recognise the limitations of deterministic planning models and the role of digital decision support.

1. The economic structure of an aggregate plan

A plan normally divides the horizon into periods such as months. Let period [[M:t]] represent month [[M:t]], where [[M:t = 1,2,...,T]]. Demand is expressed in aggregate units, equivalent units or standard labour hours. The planner chooses how much to produce internally, how much to source externally, and how much inventory or unmet demand to carry between periods.

The principal decision variables are shown below.

A basic flow-balance equation is:

Here, [[M:D_t]] is demand in period [[M:t]], and all quantities are measured in equivalent product units. The equation states that available supply plus opening net inventory must equal current demand plus closing net inventory. A positive [[M:I_t]] represents physical stock; a positive [[M:B_t]] represents demand postponed to a later period.

The equation assumes that inventory and backorders are measured consistently, that production is available in the period in which it is recorded, and that yield losses have either been incorporated into production requirements or are negligible. If production has a non-zero lead time, the relevant production variable must be shifted to the period of availability.

A common non-negative formulation uses net inventory [[M:N_t]]:

This is compact but can conceal the simultaneous holding of inventory and backorders. In most operational settings, the stronger conditions [[M:I_t ≥ 0]], [[M:B_t ≥ 0]] and a policy preventing both from being positive in the same period are appropriate.

2. Components of aggregate planning cost

2.1 Regular-time production cost

Regular-time cost includes direct labour, variable manufacturing overhead, energy and other costs that vary with internal output. If [[M:c^R_t]] is the regular-time unit cost, the cost is:

The cost is variable only if the model treats regular capacity as flexible. Salaried labour, leased facilities and other fixed costs should not be incorrectly multiplied by output. A more precise model separates fixed period costs from variable production costs.

2.2 Overtime cost

Overtime provides short-term capacity without permanently increasing the workforce. If [[M:c^O_t]] is the overtime unit cost:

Usually [[M:c^O_t > c^R_t]], because overtime involves premium wages, fatigue, supervision and possibly lower quality. Overtime capacity is limited:

where [[M:\bar{O}_t]] is the maximum feasible overtime output. Capacity limits should reflect legal working-hour restrictions, equipment availability and safe operating conditions, not merely managerial preference.

2.3 Hiring and layoff costs

Hiring cost includes recruitment, screening, onboarding, training and reduced initial productivity. Layoff cost may include severance, legal obligations, administration and loss of organisational knowledge. Workforce balance can be written as:

with [[M:H_t ≥ 0]] and [[M:L_t ≥ 0]]. Hiring and layoff cost is:

If each employee supplies [[M:a_t]] regular production hours and each unit requires [[M:r_t]] labour hours, regular capacity is:

This expression assumes all employees are equally productive, labour hours are freely assignable to the product family, and absenteeism, learning and maintenance losses have been included in [[M:a_t]]. In practice, effective capacity may be better represented by [[M:a_t(1-α_t)]], where [[M:α_t]] captures expected non-productive time.

2.4 Inventory holding cost

Inventory holding cost compensates for capital tied up in stock, storage, insurance, handling, deterioration, obsolescence and damage. If [[M:c^I_t]] is the holding cost per unit carried from period [[M:t]] to the next period:

The definition of the period-end inventory must be explicit. If a firm values average inventory, a more suitable approximation is:

The first formulation is appropriate when the charge is based on closing stock; the second is appropriate when stock changes approximately linearly during the period. Mixing the two conventions causes systematic cost errors.

If annual holding rate [[M:h]] is applied to unit value [[M:v]], and the period is a fraction [[M:m]] of a year, then:

For example, a 24% annual rate on a $50 unit held for one month gives [[M:0.24 × 50 ÷ 12 = $1.00]] per unit-month, assuming a uniform annual rate and no additional storage charges.

2.5 Shortage and backorder cost

A shortage occurs when demand cannot be met immediately. Backorders are explicitly promised for later fulfilment; lost sales are demand that disappears. Their economic consequences differ. If [[M:c^B_t]] is the cost per backordered unit per period:

Backorder cost may include expediting, customer compensation, administrative work, lost goodwill and reduced future demand. A lost-sales model instead uses a penalty for unmet demand and does not carry the quantity into subsequent inventory balance in the same way.

The shortage penalty should not be interpreted as an exact market value. It is often a managerial estimate of the incremental consequence of poor service. A low penalty can make the mathematical solution attractive but commercially unacceptable.

2.6 Subcontracting cost

Subcontracting provides capacity flexibility without hiring. If [[M:c^S_t]] is the external unit cost:

Subcontracting may also involve qualification, transport, inspection, intellectual-property protection, minimum order quantities and supplier risk. Thus, the effective cost can be higher than the quoted purchase price. A capacity limit is often required:

The make-or-buy comparison must include avoidable internal cost, quality and lead-time differences. Comparing a supplier price only with direct labour cost is usually misleading when internal capacity is available or fixed costs are unavoidable.

2.7 Idle-capacity cost

Idle capacity is the unused portion of an available resource. If available regular capacity is [[M:\bar{P}_t]] and actual regular production is [[M:P_t]], idle capacity is:

There are two distinct interpretations. Under a short-run variable-cost model, idle capacity may have zero incremental cost because fixed resources are already paid for. Under a managerial capacity-cost model, idle time is assigned a cost [[M:c^Q_t]] to represent lost contribution, labour underutilisation or opportunity cost:

The second treatment is useful for comparing alternative uses of scarce capacity, but it should not be confused with cash cost. Double counting occurs if the model includes both full fixed labour cost and a separate idle-time charge without a clear accounting rationale.

2.8 Changeover and setup cost

Changeovers consume time and may require cleaning, tooling, calibration, first-article inspection and material loss. If [[M:C_t]] is the number of changeovers and [[M:c^C_t]] is the cost per changeover:

A changeover also consumes capacity. If a setup takes [[M:s_j]] hours for product family [[M:j]], then total available production time must satisfy:

Here, [[M:x_{jt}]] is production of family [[M:j]], [[M:r_{jt}]] is labour or machine time per unit, and [[M:y_{jt}]] is a binary indicator equal to 1 when a setup occurs. Setup decisions make the model more realistic but often transform a simple linear program into a mixed-integer program.

3. The aggregate cost objective

A general deterministic objective is to minimise total relevant cost:

where [[M:F]] represents applicable fixed period costs. Expanded:

The objective is meaningful only when all terms use the same currency and planning-period basis. Costs should be adjusted for inflation where the horizon is long. If financial discounting is material, future cost may be converted to present value:

where [[M:i]] is the discount rate per planning period. Discounting is generally less important over a short operational horizon than feasibility and service, but it can matter in capital-intensive or long-horizon decisions.

4. Capacity, service and policy constraints

Cost minimisation must be bounded by operational policy. Common constraints include:

• workforce bounds: [[M:\underline{W}_t ≤ W_t ≤ \overline{W}_t]];
• regular capacity: [[M:P_t ≤ \bar{P}_t(W_t)]];
• overtime limits: [[M:O_t ≤ \bar{O}_t]];
• subcontracting limits: [[M:S_t ≤ \bar{S}_t]];
• inventory limits: [[M:0 ≤ I_t ≤ \bar{I}_t]];
• service limits: [[M:B_t ≤ \bar{B}_t]];
• terminal inventory requirement: [[M:I_T ≥ I^{target}]];
• terminal workforce requirement: [[M:W_T ≥ W^{target}]].

A terminal condition prevents the model from postponing cost beyond the horizon. For example, ending with large backorders may make the plan appear inexpensive even though the organisation begins the next horizon in crisis. Similarly, a plan that lays off nearly everyone in the final month may shift adjustment cost to the next plan.

Service can be represented by a maximum backorder, a fill-rate requirement or a lost-sales constraint. Fill rate is commonly defined as:

This measure must be defined carefully because cycle-service level, line fill rate and order fill rate are not identical. A model should use the metric that corresponds to the customer promise.

5. Major planning strategies and their trade-offs

Level strategy

A level strategy maintains a relatively stable workforce and production rate. Demand variation is absorbed through inventory, overtime, subcontracting or backorders. Advantages include workforce stability, predictable learning and lower hiring and layoff activity. Disadvantages include inventory carrying cost, obsolescence and possible shortages during demand peaks.

Chase strategy

A chase strategy varies production approximately with demand. Workforce or working time is adjusted to match requirements. It can reduce finished-goods inventory, but hiring, training, layoffs, morale and capability loss may be substantial. It is more feasible where labour is flexible and skills are readily available.

Mixed strategy

Most practical plans are mixed. The firm preserves a core workforce, uses inventory within a target range, employs limited overtime and subcontracts exceptional peaks. Mixed strategies are often superior because they recognise that each adjustment mechanism has a different cost, speed and risk.

The economic logic can be expressed as marginal comparison. If one additional unit of peak-period demand can be covered by overtime, subcontracting or inventory, the planner compares the full incremental costs, including service and capacity effects. The cheapest direct cost is not automatically the best option if it consumes scarce capacity or increases future risk.

6. Worked example: a two-period plan

The following illustrative example uses aggregate units. A firm begins with 100 units of inventory and no backorders. Demand is 600 units in month 1 and 800 units in month 2. Regular capacity is 650 units per month at $20 per unit. Overtime capacity is 150 units per month at $30 per unit. Subcontracting is available at $34 per unit. Inventory holding cost is $2 per unit per month, and backorder cost is $12 per unit per month. The firm requires ending inventory of 100 units and permits no ending backorders.

Consider the following feasible plan:

The month 1 balance is [[M:100 + 600 - 600 = 100]]. The month 2 balance is [[M:100 + 650 + 50 - 800 = 0]]. This plan does not meet the required ending inventory of 100 units, so it is infeasible under the stated terminal policy. The example demonstrates why terminal conditions must be checked rather than assumed.

A feasible alternative is:

Month 2 balance is [[M:100 + 650 + 50 + 100 - 800 = 100]]. Production cost is:

• regular time: [[M:(600 + 650) × $20 = $25,000]];
• overtime: [[M:50 × $30 = $1,500]];
• subcontracting: [[M:100 × $34 = $3,400]];
• holding: [[M:100 × $2 × 2 = $400]], if closing inventory is charged in both months;
• total relevant cost: [[M:$25,000 + $1,500 + $3,400 + $400 = $30,300]].

The calculation assumes that inventory of 100 units is held at the end of both months and that no hiring, layoff, idle-capacity or changeover cost applies. If month 2 regular production is reduced and subcontracting increases, direct cost rises because subcontracting costs more than regular production. However, such a decision could be justified if regular capacity is needed for another product, if overtime is restricted, or if supplier capacity is strategically valuable.

7. An illustrative case: Northstar Components

Background

Northstar Components, a fictional manufacturer of industrial control modules, supplies three related products assembled on a common line. Demand is seasonal: customer orders rise sharply in the third quarter. The company has 42 trained operators, a regular capacity of 7,200 equivalent units per month, and a maximum overtime capacity of 1,000 units. Qualified subcontractors can provide up to 1,500 units monthly, but their price is 18% above Northstar’s regular variable manufacturing cost.

Management historically used a chase strategy. During the seasonal peak, it hired temporary workers and then released them within four months. An internal review identified training delays, quality escapes and employee dissatisfaction. The operations director requested a cost-and-service comparison of three strategies.

Problem

The decision team had to satisfy a two-month peak demand of 16,800 units while maintaining a minimum monthly service level of 98%. The initial inventory was 1,000 units. The company also required at least 1,000 units at the end of the horizon to protect the next month’s customer commitments.

Analysis

The team first separated relevant costs from accounting allocations. Regular variable cost was $24 per unit, overtime cost was $36, subcontracting cost was $43, inventory holding cost was $1.50 per unit-month, and backorder cost was estimated at $20 per unit-month. Hiring cost was $4,500 per operator and layoff cost was $3,000 per operator. A quality audit estimated that temporary labour increased rework by 2.5 percentage points during the first month of assignment.

Three alternatives were modelled:

1. Chase: hire enough operators to cover the peak and lay them off afterward.
2. Level: retain the existing workforce, use inventory and subcontracting, and avoid layoffs.
3. Mixed: retain the core workforce, use overtime within safe limits, subcontract the residual peak and build inventory before the peak.

The mixed plan had a slightly higher direct conversion cost than the nominal chase plan, but it reduced expected rework, protected service and avoided repeated recruitment. The model also included a penalty for exceeding the subcontractor’s qualified capacity and a maximum backorder equal to 2% of monthly demand.

Approach and solution

Northstar selected the mixed plan. It produced at a stable regular rate before the peak, built 1,200 units of inventory, scheduled overtime for the first peak month, and subcontracted only the residual requirement in the second peak month. No layoffs were scheduled. Changeovers were reduced by grouping products into longer production campaigns, although this increased average finished-goods inventory slightly.

The plan was implemented with a rolling horizon. Each month, the first period was frozen except for emergencies, while later periods were re-optimised using updated demand forecasts and supplier confirmation. A dashboard reported total cost, fill rate, overtime hours, inventory age, backorders, absenteeism and supplier defects.

Outcome

The plan achieved the required service level and maintained the terminal inventory target. Its direct cost was not the lowest under the original deterministic assumptions, but the total realised cost was lower than the chase strategy after rework, recruitment and expedited freight were included. Workforce turnover declined, and the supplier relationship became more predictable because orders were released earlier.

Learning points

The case illustrates five principles. First, aggregate plans should include strategic and quality consequences, not only conversion cost. Second, a small amount of inventory can be economical when it prevents expensive peak adjustments. Third, subcontracting is valuable as a flexibility option but creates dependency and quality exposure. Fourth, changeover reduction can improve both capacity and cost. Finally, rolling-horizon control is essential when forecasts and supplier conditions change.

8. Managerial interpretation of cost trade-offs

Cost versus service

Reducing inventory and overtime may lower cost while increasing late deliveries. The appropriate decision depends on customer segmentation and contractual commitments. A premium customer may justify a high service target, while a low-margin, highly predictable order may tolerate a longer lead time.

Cost versus workforce stability

Hiring and layoffs provide flexibility but can destroy firm-specific knowledge. Stable employment may reduce visible adjustment cost but create idle time in low-demand periods. The correct comparison includes learning curves, morale, safety, absenteeism and future recruitment difficulty.

Cost versus capacity risk

Subcontracting shifts production risk to a supplier. It can protect the firm from irreversible investment, but supplier failure, confidentiality loss or quality variation may be unacceptable for critical components. Dual sourcing and qualification audits can reduce, but not eliminate, this exposure.

Cost versus environmental and social performance

Overproduction increases material use and possible obsolescence. Excessive overtime can create fatigue and safety risks. Layoffs can have serious social consequences. Modern aggregate planning therefore increasingly uses multi-objective analysis, reporting financial cost alongside emissions, energy, workforce and service indicators.

9. Modelling methods and technology

A small aggregate model can be solved with a spreadsheet, provided formulas, units and constraints are auditable. Linear programming is appropriate when costs and capacity relationships are approximately linear. Mixed-integer programming is needed for setup decisions, minimum run lengths, fixed subcontracting charges or workforce start-up choices. Simulation is valuable when demand, processing time, yield or supplier reliability is stochastic.

Artificial intelligence can improve demand sensing, anomaly detection and scenario generation, but it does not remove the need for managerial assumptions. A machine-learning forecast may identify patterns while remaining vulnerable to structural breaks, promotions or data bias. Governance should require traceable inputs, versioned assumptions, human review of exceptional recommendations and monitoring for systematic service discrimination among customer groups.

A robust workflow is:

1. define the product family and unit of aggregation;
2. validate demand, capacity and workforce data;
3. specify cost definitions and terminal conditions;
4. create a base deterministic model;
5. test alternative strategies and sensitivity ranges;
6. stress-test demand, yield, supplier and labour shocks;
7. obtain cross-functional approval;
8. execute through a rolling-horizon process;
9. compare actual outcomes with plan assumptions.

10. Limitations and best practices

Aggregate models conceal variation among products. Two products may have the same equivalent-unit requirement but very different materials, skills or customer priorities. Aggregation can therefore create an apparently feasible plan that is infeasible at the detailed scheduling level.

Other limitations include forecast error, uncertain productivity, non-linear learning, minimum batch sizes, imperfect supplier quality and the difficulty of assigning a monetary value to goodwill. Cost parameters should be treated as estimates and tested through sensitivity analysis. If a recommendation changes radically when backorder cost moves from $18 to $22, management should recognise that the decision is assumption-sensitive.

Best practices include:

• use capacity in hours as well as units where product mix matters;
• distinguish backorders from lost sales;
• include beginning and ending inventory explicitly;
• prevent terminal-period cost shifting;
• separate cash cost, accounting allocation and opportunity cost;
• include setup time in capacity, not only setup cost;
• evaluate quality, safety and legal limits on overtime;
• use rolling horizons and scenario analysis;
• reconcile the aggregate plan with master scheduling and material planning;
• document who owns each assumption.

Chapter Summary

Aggregate planning balances demand with regular production, overtime, subcontracting, inventory and workforce decisions over a medium-term horizon. Its core accounting identity links opening inventory and backorders, current supply and demand, and closing inventory and backorders. A credible model includes regular-time, overtime, hiring, layoff, holding, shortage, subcontracting, idle-capacity and changeover costs where they are economically relevant.

The lowest direct cost is not necessarily the best plan. Level strategies promote workforce stability but may require inventory; chase strategies reduce inventory but increase workforce adjustment; mixed strategies combine mechanisms according to their marginal costs, limits and risks. Capacity, service and terminal conditions are essential constraints. Practical decisions should also consider quality, safety, supplier reliability, environmental effects and employee welfare.

Linear programming, mixed-integer programming, simulation and digital forecasting tools support analysis, but their results depend on valid data and transparent assumptions. A rolling-horizon process, scenario testing and cross-functional governance convert a static cost model into a useful management system.

Key Terms

Aggregate planning: Medium-term coordination of demand, capacity, workforce and inventory for product families.

Backorder: Customer demand intentionally postponed for later fulfilment.

Changeover cost: Cost associated with switching equipment, tools, materials or product families.

Chase strategy: Strategy that adjusts production or workforce to follow demand.

Fill rate: Proportion of requested units supplied immediately.

Idle capacity: Available production capability that is not used.

Inventory holding cost: Cost of storing and financing inventory over time.

Level strategy: Strategy that maintains relatively stable workforce and production.

Lost sales: Demand that is not fulfilled and does not remain as a future requirement.

Mixed strategy: Aggregate plan combining several adjustment mechanisms.

Net inventory: Inventory minus backorders.

Overtime: Production performed beyond regular working capacity.

Regular-time production: Output produced within normal scheduled capacity.

Service constraint: Requirement limiting shortages or specifying a customer-service target.

Shortage cost: Economic penalty associated with unmet or delayed demand.

Subcontracting: Obtaining production from an external supplier.

Terminal condition: Requirement imposed at the end of the planning horizon.

Workforce balance: Relationship among opening workforce, hires, layoffs and closing workforce.

Review Questions

Short Answer Questions

1. Why must inventory and backorders be distinguished in an aggregate planning model?
2. Identify three cost components associated with changing workforce size.
3. What is the difference between idle-capacity cost and regular fixed cost?
4. Why are terminal inventory and workforce conditions important?
5. State one advantage and one disadvantage of subcontracting.

Long Answer Questions

1. Explain the cost and service trade-offs among level, chase and mixed aggregate planning strategies.
2. Develop and discuss a total-cost model incorporating regular production, overtime, subcontracting, inventory and backorders.
3. Explain how changeover decisions affect both cost and available capacity.
4. Critically evaluate the use of backorder penalties as a representation of customer-service consequences.
5. Discuss how workforce stability, quality and safety should influence an aggregate plan beyond direct financial cost.

Analytical / Application Questions

1. A plant begins with 200 units of inventory. Demand is 700 units in period 1 and 900 units in period 2. Regular production is limited to 650 units per period, overtime to 150 units per period, and subcontracting is unlimited. Construct a feasible plan that ends with 100 units of inventory, and identify the quantities obtained from each source.
2. A manager proposes reducing inventory holding cost by using more backorders. Design a sensitivity analysis showing which cost, service and terminal assumptions should be varied before accepting the proposal.
3. A product family requires different machine times and has four product changeovers in a month. Explain why a unit-only aggregate model may be infeasible, and propose the additional data and constraints needed for a mixed-integer planning model.

Multiple Choice Questions with Answers

Question 1

Which equation correctly represents aggregate inventory and backorder balance?

A. [[M:I_t = D_t + P_t]]

B. [[M:I_{t-1} - B_{t-1} + P_t + O_t + S_t - D_t = I_t - B_t]]

C. [[M:W_t = W_{t-1} - H_t - L_t]]

D. [[M:P_t = W_t + D_t]]

Correct Answer: B. [[M:I_{t-1} - B_{t-1} + P_t + O_t + S_t - D_t = I_t - B_t]]

Question 2

Which cost most directly represents recruitment, screening and onboarding?

A. Inventory holding cost

B. Changeover cost

C. Hiring cost

D. Backorder cost

Correct Answer: C. Hiring cost

Question 3

A firm has regular capacity of 650 units and produces 500 units in a period. Under the definition in the chapter, idle capacity is:

A. 150 units

B. 500 units

C. 650 units

D. 1,150 units

Correct Answer: A. 150 units

Question 4

Which strategy generally maintains a relatively stable workforce and uses inventory to absorb demand variation?

A. Chase strategy

B. Level strategy

C. Lost-sales strategy

D. Subcontract-only strategy

Correct Answer: B. Level strategy

Question 5

If annual holding rate is 24% and unit value is $50, the holding cost for one month is:

A. $0.24

B. $0.50

C. $1.00

D. $12.00

Correct Answer: C. $1.00

Question 6

Why must setup time be included in a capacity model?

A. It always eliminates inventory cost

B. It consumes available production time

C. It converts backorders into lost sales

D. It guarantees lower overtime cost

Correct Answer: B. It consumes available production time

Question 7

Which condition prevents a plan from postponing cost beyond the planning horizon?

A. A terminal inventory or workforce requirement

B. A higher regular-time cost

C. A lower demand forecast

D. Removing subcontracting limits

Correct Answer: A. A terminal inventory or workforce requirement

Question 8

What is a principal risk of relying heavily on subcontracting?

A. Guaranteed elimination of supplier risk

B. Supplier quality or capacity dependence

C. Automatic reduction in lead time

D. Inability to use any inventory

Correct Answer: B. Supplier quality or capacity dependence

Question 9

A two-period plan starts with 100 units, produces 600 units, and faces demand of 600 units in period 1. Ending inventory is:

A. 0 units

B. 50 units

C. 100 units

D. 700 units

Correct Answer: C. 100 units

Question 10

Which method is especially appropriate when fixed setup decisions and minimum run lengths are modelled?

A. Mixed-integer programming

B. Simple averaging only

C. Unconstrained forecasting

D. Cycle-service calculation alone

Correct Answer: A. Mixed-integer programming