Econ Papers Daily / 期刊论文
社会比较反馈与居民节约用电：来自墨西哥的实地实验证据
Social comparison feedback and residential electricity conservation: Field experimental evidence from Mexico
摘要
Residential electricity demand in emerging economies is rising faster than in advanced economies, driven by income growth, urbanization, and warming temperatures, yet the price reforms typically used to curb demand are often politically infeasible where electricity is subsidized. We test whether a low-cost, non-price alternative can reduce residential electricity consumption in this setting. In two randomized field experiments in Mérida and Cancún, fast-growing cities on Mexico’s Yucatán Peninsula, households received a hand-delivered flyer that compared their electricity use with that of their neighbors and listed conservation tips. Using a difference-in-differences design, we find that in the first billing period after delivery, treated households reduced consumption by 11.3% in Cancún and 9.3% in Mérida, and their bills by 18.9% and 11.0%. These estimates are robust to conditioning on baseline consumption, to using the consumption figures originally recorded in the field, and to bounds for selective attrition. In both cities, households with higher baseline consumption reduced their consumption most, in absolute and percentage terms, suggesting that utilities could target such interventions using billing data they already hold; in Cancún, households told they consumed more than their neighbors also responded more strongly. Because our samples are small and our horizon short, we interpret these effects, which lie well above the 1–2% benchmark from large U.S. trials, as a likely upper bound on what a scaled program would achieve. We argue that the scarcity of comparable evidence from the Global South is itself a policy-relevant gap.
中文摘要
新兴经济体的居民用电需求增速快于发达经济体，其驱动因素包括收入增长、城市化和气温升高，然而在电力享有补贴的情况下，通常用于抑制需求的价格改革往往在政治上不可行。我们检验在这一情境中，一种低成本、非价格型的替代方案能否降低居民用电量。在墨西哥尤卡坦半岛快速扩张的城市梅里达和坎昆开展的两项随机实地实验中，住户收到手工递送的单页传单，其中将其用电量与邻居的用电量进行比较，并列出节能建议。采用双重差分设计，我们发现，在传单递送后的第一个计费周期内，处理组住户在坎昆的用电量降低了11.3%，在梅里达降低了9.3%，其电费分别降低了18.9%和11.0%。这些估计在控制基线用电量、使用实地最初记录的用电量数据以及针对选择性流失设置界限的条件下均保持稳健。在两座城市中，基线用电量较高的住户用电量下降幅度最大（无论以绝对值还是百分比衡量），这表明公用事业公司可利用其已持有的计费数据来定向实施此类干预；在坎昆，被告知自身用电量高于邻居的住户反应也更强。由于我们的样本较小且观察期较短，我们将这些效应——其远高于美国大型试验1–2%的基准——解读为规模化项目可能实现效果的上限。我们认为，来自全球南方的可比证据稀缺本身就是一个与政策相关的空白。