**1. Introduction**

Improving sustainability on the production and consumption sides of product life cycles has proven to be critical in reducing the carbon footprint and combating global warming [1]. For this reason, one of the United Nations Sustainable Development Goals (i.e., Goal #12) explicitly addresses the problems associated with unsustainable production and consumption (https://www.un.org/sustainabledevelopment/sustainable-developmentgoals/ (accessed on 22 December 2021). Many manufacturers shift production to low-cost and distant countries to benefit from low production costs, but the long production and shipping lead times between production and the market bases contribute to significant amounts of excess inventory [2] that risk going to waste in retail stores without ever reaching consumers. The cost of excess inventory in the retail industry was estimated to be USD 471 billion in 2014 [3]. In other words, the Earth's resources to a value of USD 471 billion are wasted in producing goods that are never sold, and hence never used, by any consumer.

Let us consider the apparel industry, which is responsible for 8–10% of global carbon emissions [4]. The industry is dominated by strong brands that outsource production to contract manufacturers in offshore countries that rely on coal-fueled power plants. These contract manufacturers sometimes even outsource production to yet other countries to further reduce production costs and increase their capacity to fulfill increasing global demand [5]. These offshoring waves have severe effects on the environment. The industry is reported to be responsible for around 35% of oceanic microplastic pollution, 20% of industrial water pollution, and more than 8% of global carbon emissions [4]. Despite this environmental destruction, for 30–40% of clothes produced, there is no customer demand [6], resulting in a loss of profit for the apparel brands. Therefore, 30–40% of the environmental disaster could be eliminated by avoiding holding excess inventory, which is also appealing for retailers because it helps them increase their profits.

Operational flexibility has been proposed by scholars as an effective method for minimizing mismatches between supply and demand under demand uncertainty [2]. If demand exceeds supply, companies incur the opportunity cost of losing the demand. If demand falls short of supply, companies end up with excess inventory and incur inventory holding costs. In addition to the negative impact on profits, excess inventory has a catastrophic impact on the environment due to the carbon emissions and pollution that arise during the production and logistics operations for goods that are not even demanded by customers. It is therefore important to conduct a comprehensive study of operational and environmental trade-offs arising from the interaction between different supply chain processes such as procurement and inventory management [7]. In the extant literature, the merits of operational flexibility are quantified from the perspective of its impact on profits [2,8–11]. However, its benefits for environmental sustainability have not been addressed yet. In this research, we aim to fill this gap in the literature by addressing the following two questions:


We consider three different operational-flexibility strategies. The first is lead-time reduction, which can be achieved by localizing production near the market bases. Leadtime reduction allows a buyer to postpone ordering decisions until credible information from the market about the final demand has been collected. Therefore, decision makers base their decisions on accurate demand forecasts and hence are able to reduce supply–demand mismatches [2]. Second, we analyze quantity flexibility whereby an offshore supplier offers the buyer flexibility to update the initial order quantity, within some limits, after the buyer has improved its demand forecasts [10]. Compared with lead-time reduction, quantity flexibility does not require the localization of production near the market bases. Finally, we consider multiple sourcing, in which a buyer employs a domestic supplier and an offshore supplier to exploit the market responsiveness of the domestic supplier and the cost efficiency of the offshore supplier at the same time [9]. Although it has been well established in the extant literature that these three strategies are highly effective in reducing supply–demand mismatches, their impact on reducing waste is not well known. We assume a profit-oriented buyer who aims to maximize profit and employs operational-flexibility strategies just to reduce mismatch costs. Based on the profit-maximizing decisions of the buyer, we quantify the secondary positive impacts of the operational-flexibility strategies on environmental sustainability.

Following [12], we use a multiplicative demand process to model the evolutionary dynamics of demand uncertainty. Then, we quantify the impact of key modeling parameters for each operational-flexibility strategy on the waste ratio, which is measured as the ratio of excess inventory when a certain operational-flexibility strategy is employed to the amount when an offshore supplier is utilized without any operational flexibility. Suppose, for example, the expected excess inventory is 100 units if a buyer sources products from an offshore supplier. Then, the supplier offers the buyer quantity flexibility, helping the buyer reduce the expected excess inventory to 60 units. For the quantity-flexibility strategy employed, the waste ratio obtained is 60/100 = 60%. Our results show that the lead-time reduction strategy has the maximum capability to reduce waste in the sourcing process of buyers, followed by the quantity-flexibility and multiple-sourcing strategies, in order. Therefore, operational-flexibility strategies that rely on the localization of production are key to reducing waste and improving environmental sustainability at source.

We organize the remainder of the paper as follows. In Section 2, we position our research by reviewing the extant literature on circular operations management and operational flexibility. We present the model preliminaries in Section 3. Then, we analyze each

operational strategy and present some numerical examples in Section 4, where we also discuss the environmental implications further in Section 5. Finally, we provide concluding remarks and envision future research directions in Section 6.
