Elysia Slakie’s scientific contributions

a1 Increasing farm productivity is widely regarded as a prerequisite for improving rural livelihoods and supporting development in low-income countries. Public and private institution mission statements and budgets reflect this priority: the U.N. Food and Agriculture Organization (FAO) mandate is to " improve agricultural productivity, better the lives of rural populations and contribute to the growth of the world economy " (1); fully 75% of World Bank agricultural lending goes to " increasing agricultural productivity " (2); the U.S. Agency for International Development identifies increased productivity as a key to " inclusive agriculture-led growth " (3); and the Bill & Melinda Gates Foundation invests " to reduce hunger and poverty for millions of farming families in Sub-Saharan Africa and South Asia by increasing agricultural productivity in a sustainable way " (emphasis added) (4). Toward productivity-related goals, governments, non-profits and others have invested billions to raise crop yields (5). But despite good intentions, yield-based crop output measures – used ubiquitously to identify successful interventions and target needy households – are a poor proxy for smallholder farm productivity. While exact measurement methods are rarely specified, yield is most often defined as quantity of harvested crop divided by area harvested, typically expressed in kilograms or metric tons per hectare: The merits of this indicator include its ease of calculation, intuitive interpretation, widespread acceptance, and the availability of time-series data allowing for comparisons of yield estimates (1, 5). Particularly in developing country contexts, however, this common yield measure is also known to be imperfect – both in terms of the crude data used to estimate yield, and also in terms of the validity of the yield measure itself as an indicator of productivity. Meeting mission statements of inclusive, sustainable smallholder farmer productivity growth, we argue, requires recognizing that the most common yield measures appear to carry systematic biases, and true productivity measures must include all relevant farm outputs and inputs. YIELD (MIS)MEASUREMENT. Historically, published crop yield estimates were determined by experts who would harvest and weigh crops from controlled plots of known area. But even under experimental settings researchers identified at least three distinct yield measures: (i) biological (gross) yield, defined as yield before any harvest or post-harvest losses; (ii) harvested yield, defined as yield less any losses during harvest-related processing; and (iii) economic yield, defined as yield less harvest or post-harvest losses (such as spoilage) (6). Thus even early on the question " what is yield? " was not easy to answer – was it food growing on the field (i), food coming off the field (ii), or food available to be eaten or sold by the farmer (iii)? For little reason other than ease of measurement, agronomists chose (ii). Calculations of smallholder yields are further complicated by irregular plot shapes, non-planted areas due to stumps, anthills/termite mounds and other obstructions, and harvesting extended over long periods (as with root and tuber crops stored in the ground until needed) that can make " area harvested " difficult to estimate (6). Measuring yield as " kilograms of crop X produced per hectare harvested " also ignores the very real possibility that smallholder farmers might experience crop failure between planting and harvesting. In other words, in many cases crop area harvested may be smaller than crop area planted due to poor germination, damage from pests or disease, floods, insufficient farm labor, or limited market