In 2012, New York City considered implementing a policy that would prevent restaurants from selling sugary drinks in excess of 16 oz. In our 2017 article, “Psychologically Informed Implementations of Sugary-Drink Portion Limits,” we considered two plausible ways that restaurants might comply with such a policy and the effects of both on people’s propensity to buy and consume sugary drinks ( John, Donnelly, & Roberto, 2017). In three experiments, we assessed the impact of providing free refills (e.g., offering a regulation-size, 16-oz cup with unlimited refills), which led people to consume more, especially when the drinks were served by a waiter. In a final experiment, we tested bundling (i.e., dividing the contents of a 32-oz cup into two regulationsize, 16-oz cups); this led people to purchase fewer ounces of sugary drinks but did not affect the amount they consumed. In each laboratory experiment, participants made real ordering decisions, paid for their beverages, and consumed those beverages. The findings from the bundling experiment ran counter to Wilson, StolarzFantino, and Fantino’s (2013) findings on beverage bundling with 100 undergraduates making hypothetical purchasing decisions. In that study, they concluded that “restricting larger-sized drinks may have the unintended consequence of increasing soda consumption rather than decreasing it” (Abstract) if companies respond to the policy by offering beverage bundles. Wilson and Stolarz-Fantino (2018) offer two primary critiques, one specific to our research and one more broadly applicable to policy-relevant social science. First, they argue that our laboratory context was not sufficiently representative of the marketplace; in particular, we used linear pricing for beverages (instead of restaurants’ more common nonlinear pricing structure), and therefore our results cannot inform our understanding of how this policy might play out in the real world. Second, they argue that “for their research to have policy implications, John et al. would need to demonstrate that their menus are profit maximizing” (p. 1376). We address both critiques in turn below.

中文翻译：

---

使用行为科学为限制含糖饮料部分的政策提供信息：回复 Wilson 和 Stolarz-Fantino (2018)

2012 年，纽约市考虑实施一项政策，禁止餐馆销售超过 16 盎司的含糖饮料。在我们 2017 年的文章“含糖饮料部分限制的心理知情实施”中，我们考虑了餐厅可能遵守此类政策的两种可能方式，以及两者对人们购买和消费含糖饮料倾向的影响（John、Donnelly 和罗伯托，2017 年）。在三个实验中，我们评估了提供免费续杯的影响（例如，提供一个规定尺寸的 16 盎司杯子，无限续杯），这会导致人们消费更多，尤其是当饮料由服务员端上时。在最后的实验中，我们测试了捆绑（即，将 32 盎司杯子的内容物分成两个规定尺寸的 16 盎司杯子）；这导致人们购买较少盎司的含糖饮料，但并未影响他们的消费量。在每个实验室实验中，参与者做出真正的订购决定，为他们的饮料付款，并消费这些饮料。捆绑实验的结果与 Wilson、StolarzFantino 和 Fantino (2013) 关于 100 名大学生做出假设购买决定的饮料捆绑的发现背道而驰。在该研究中，他们得出结论，如果公司通过提供饮料捆绑来响应该政策，“限制大容量饮料可能会导致增加苏打水的消费量而不是减少它的意外后果”（摘要）。Wilson 和 Stolarz-Fantino (2018) 提供了两种主要批评，一种专门针对我们的研究，另一种更广泛地适用于与政策相关的社会科学。第一的，他们认为我们的实验室环境不足以代表市场；特别是，我们对饮料使用了线性定价（而不是餐厅更常见的非线性定价结构），因此我们的结果无法让我们了解该政策在现实世界中可能如何发挥作用。其次，他们认为“对于他们的研究具有政策意义，约翰等人。需要证明他们的菜单是利润最大化的”（第 1376 页）。我们在下面依次处理这两种批评。他们争辩说，“对于他们的研究具有政策意义，约翰等人。需要证明他们的菜单是利润最大化的”（第 1376 页）。我们在下面依次讨论这两种批评。他们争辩说，“对于他们的研究具有政策意义，约翰等人。需要证明他们的菜单是利润最大化的”（第 1376 页）。我们在下面依次讨论这两种批评。