decline in childhood measles, mumps and rubella (MMR) vaccinationfollowing communication of a singleflawed study in 1998, and themore recent rise in a broader anti-vaccination movement (Hussainet al., 2018). Genetically modified organism (GMO) food sources are anexample within agriculture. Despite scientific consensus suggesting thatthere is no greater risk to the environment or consumer health inconsuming GMO food than conventional food, GMO food is not globallyaccepted or utilised (Yang and Chen, 2016;Scott et al., 2018). Whilesome may oppose GMO with concerns that large companies may gainpower and influence over producers (Dibden et al., 2013); the debate isoften impervious to science and instead guided by emotional andmoralising rhetoric (Hielscher et al., 2016;Scott et al., 2018).Perceptions and opinions of the global livestock sector are not onlydiverse but are also becoming increasingly polarised (Busch and Spiller,2018). For instance, based on evidence concerning negative aspects oflivestock, the proportion of the population in high-income countriesfollowing vegan diets is rapidly increasing (Gill et al., 2018). With thesame evidence, some commentators call for a livestock-free world(Monbiot, 2017). Such a stance ignores the role livestock productionplays in supporting the livelihoods of large populations in low- andmiddle-income countries (LMICs) (Herrero et al., 2013a). By improvingcommunication and interpretation of livestock evidence, including bothnegative and positive impacts of the livestock sector, today's polarisedviews might be replaced by a more constructive dialogue concerninglivestock's role in humanity's future.Livestock Data for Decisions (LD4D) is a community of practice,with members from academia, NGOs, donor agencies and industry.LD4D's aim is to‘drive informed livestock decision-making throughbetter use of existing data and analyses’(LD4D, 2019). This multi-dis-ciplinary community, seeing that livestock sector‘facts’are regularlycommunicated without a full understanding of their provenance andvalidity, initiated a Livestock Fact Check project to identify and in-vestigate popular‘facts’relating to important aspects of livestock sys-tems. This project encouraged dialogue to help ensure livestock pro-duction discussions and decisions are well informed, facts appropriatelyinterpreted, and gaps in knowledge and assumptions recognised. Thispaper summarises thefindings of the project's investigations.2. Investigating popular livestock‘facts’The livestock‘facts’investigated by LD4D refer both to quantitativeevidence, such as a specific percentage or quantity, and to qualitativeevidence, such as a broad perception or understanding. The’facts' to beinvestigated were identified by a multidisciplinary working group re-presenting the broader multi-discplinary LD4D community of practice.To help promote constructive community dialogue, effortswere madeto select‘facts’that represented key aspects of livestock production (e.g.economy, environment, livelihoods and health). While the methodol-ogies used to investigate each‘fact’were variable, efforts were made inevery case to identify the origins of the‘facts’through interrogations ofthe relevant literature. The‘facts’investigated are summarised in thefollowing sections.2.1. Livestock supporting the livelihoods of poor peoplePublications on livestock development have regularly quoted‘onebillion’as being the global number of poor people supported by live-stock, typically to demonstrate the importance of livestock to humanity.In some instances this number is not given a reference (e.g.FAO, 2009);more commonly, a secondary source is cited (e.g.Thorne and Conroy,2017). Tracing back through publications referencing‘one billion’suggests that the original source is a study published in 1999 by Live-stock in Development (LID). The UK Department for International De-velopment (DFID) funded that study to examine‘the case for investmentin the livestock sector as a basis for reducing rural poverty’(LID, 1999). LIDcalculated that the global number of poor livestock-keepers was 987million (rounded up to one billion). For this calculation, LID used aglobal livestock-keeper agro-ecological distribution reported bySeréand Steinfeld (1996)and poverty statistics from theUnited NationsDevelopment Programme (1997). In turn, the livestock-keeper dis-tribution was based on 1991–1993 data from AGROSTAT (now FAO-STAT) and the poverty statistics used a composite poverty measurecited as“correspondence on the Gini coefficient”from the World Re-sources Institute in 1996 (no further information was available relatingto this correspondence). Consequently, the regularly quoted‘one bil-lion’statistic is based on a calculation using previous publications andstatistics, all more than 20 years old. Between 1999 and 2017, theglobal human population has grown from 6.1 to 7.6 billion (FAOSTAT,2019); growth is concentrated in LMICs and unlikely to slow in theforeseeable future (Gerland et al., 2014). Globally, there is net rural tourban migration, with a higher proportion of the population living inurban centres than in rural locations since 2007 (FAO, 2018d;FAOSTAT, 2019). Recent economic modelling suggests that for thoseremaining rural populations, particularly farmers, poverty is limiting,with economic growth suggested to be low or even negative (Castañedaet al., 2018; Laborde Debucquet and Martin, 2018). Additionally, theglobal definitionof‘poor’has changed; the World Bank's one-dollar-a-day indicator, tracking‘the share of individuals that have to live on lessthan an absolute minimum’, was adjusted three times between 1999 and2015 (Klasen et al., 2016). Today, the number of people living in‘ex-treme poverty’, below the USD-1.9-a-day poverty line, is estimated tobe 731 million (The World Bank, 2019), which invalidates the‘onebillion’poor livestock keepers statistic. There are more recent calcula-tions to suggest livestock's significance in supporting the global poor(Herrero et al., 2009;Robinson et al., 2011); however, variation inmethods and applied data mean trends cannot be easily assessed. Ac-curacy may not be of high importance in certain applications of this‘fact’. For instance, the order of magnitude could be enough to com-municate that a‘large number’of poor people depend on livestock, andtherefore livestock should not be ignored in future food security efforts(FAO, 2018c; World Economic Forum, 2019). Whereas to inform spe-cific investment and development actions (and monitor subsequentimpact) it would be useful to understand trends (requiring increasedaccuracy and a consistent methodology) in the importance of livestockat both local and global scales. Interestingly, more recent publicationsdemonstrate the need to give livestock development consideration bydrawing attention to the importance of livestock to stakeholders acrossassociated value chains, not just poor livestock keepers (WorldEconomic Forum, 2019).Fig. 1.Global trend in the use of Google search term‘Livestock’. The y-axisnumbers 0–100 represent search interest relative to the highest point on the2014–2019 chart. A value of 100 is the peak popularity for the term (whichoccurred in March 2019). A value of 50 means that the term is half as popular.Data from Google trends (Google, 2019).G.R. Salmon, et al.Global Food Security xxx (xxxx) xxxx2