Accuracy issues in FAO animal numbers

Some strate­gic de­ci­sions in an­i­mal ad­vo­cacy are in­formed by the num­bers of an­i­mals in var­i­ous coun­tries. The most widely used statis­tics about the num­bers of farmed an­i­mals come from the Food and Agri­cul­ture Or­ga­ni­za­tion of the United Na­tions (FAO) web­site FAOSTAT. In this ar­ti­cle, I provide some ex­am­ples of in­con­sis­ten­cies in FAOSTAT an­i­mal data, and fur­ther rea­sons to think that the data may some­times be in­ac­cu­rate. The main point of this ar­ti­cle is that it’s prob­a­bly worth try­ing to ver­ify FAO num­bers be­fore us­ing them to make im­por­tant de­ci­sions. In the ap­pendix, I also ex­plain some minor caveats that should be un­der­stood when in­ter­pret­ing FAO num­bers.

What rele­vant statis­tics FAO provides

For land an­i­mals, FAO pro­vides these statis­tics for each coun­try and each year:

Some of these statis­tics are vi­su­al­ized by Our World in Data and Fau­n­a­lyt­ics.

Statis­tics about fish farm­ing and wild fish catch for each coun­try, year, and species can be seen by down­load­ing the FishS­tatJ pro­gram. Some cu­mu­la­tive statis­tics are pre­sented and vi­su­al­ized in The State of World Fish­eries and Aqua­cul­ture (SOFIA) 2018 re­port.

Known issues

FAO claims that “it is not pos­si­ble to as­sess the over­all ac­cu­racy of the dataset, as the source data is largely col­lected by mem­ber coun­tries.”[1] I’d guess that some coun­tries don’t have good statis­tics and that there is sig­nifi­cant ac­tivity that is not cap­tured by gov­ern­ment statis­tics.

For ex­am­ple, page 93 in FAO’s SOFIA 2018 re­port claims that “it is rec­og­nized that the FAO cap­ture database does not in­clude all fish caught in the wild, as it omits the por­tion of the catch that is dis­carded at sea and catches from ille­gal, un­re­ported or un­reg­u­lated (IUU) fish­eries.” Pauly and Zel­ler (2016) es­ti­mate that be­cause of this is­sue, global catches be­tween 1950 and 2010 were roughly 50% higher than data re­ported to FAO sug­gest. Since web­sites like fish­count.org.uk use FAO data for their es­ti­mates, they are prob­a­bly un­der­es­ti­mat­ing wild-caught fish num­bers as well.

Another po­ten­tial prob­lem is that coun­try offi­cials may not always have in­cen­tives to re­port ac­cu­rate data. For ex­am­ple, ac­cord­ing to Yu and Abler (2014), China has been over­re­port­ing pork pro­duc­tion be­cause lo­cal offi­cials have been in­flat­ing pro­duc­tion figures to im­prove their prospects for pro­mo­tion. How­ever, Xiao et al. (2015) claims that over the last decades, the ac­cu­racy of China’s meat pro­duc­tion statis­tics have been sig­nifi­cantly im­proved as the in­cen­tives to over-re­port agri­cul­tural pro­duc­tion have grad­u­ally dis­ap­peared. I’m un­sure if there are similar prob­lems in other coun­tries. Note that when mis­takes or mis­re­port­ing is no­ticed, in some cases FAO may cor­rect its statis­tics. E.g., see FAO (2001).

Fi­nally, FAO data for some an­i­mals may be in­com­plete. For ex­am­ple, ac­cord­ing to Wald­horn (forth­com­ing), FAO does not in­clude snail pro­duc­tion figures for any Euro­pean coun­try, de­spite other sources in­di­cat­ing that snails are farmed in the con­ti­nent. Similarly, FAO data on ro­dent farm­ing only cov­ers Peru and Bo­livia, even though ro­dents are also farmed in Africa (see Maass et al. (2014)).[2]

Ex­am­ples of inconsistencies

When an­a­lyz­ing FAO statis­tics for land an­i­mals, I no­ticed some in­con­sis­ten­cies. Hence, I started com­par­ing FAO statis­tics with offi­cial statis­tics from var­i­ous coun­tries to un­der­stand what is go­ing on and no­ticed fur­ther po­ten­tial is­sues. I re­ported some of these is­sues on the FAOSTAT feed­back fo­rum and they promised to re­vise the data at the end of this year. Nev­er­the­less, there are prob­a­bly many other is­sues with the data that will re­main un­cor­rected. I wanted to doc­u­ment my rea­son­ing to show­case what kind of prob­lems can be ex­pected. Note that the ori­gin of these is­sues is un­clear. They could be due to mis­re­port­ing by coun­tries, mis­es­ti­mat­ing by FAO, or mis­com­mu­ni­ca­tion be­tween coun­tries and FAO. For brevity, I use ‘M’ for a mil­lion and ‘B’ for a billion.

Egg-lay­ing hens

Bangladesh

Ac­cord­ing to FAO statis­tics, there are 275M chick­ens in Bangladesh (based on offi­cial data) and 301M egg-lay­ing hens (FAO es­ti­mate) al­ive at any time. It makes no sense that there are more egg-lay­ing hens than chick­ens in to­tal be­cause the to­tal num­ber of chick­ens should in­clude egg-lay­ing hens.[3]

This in­con­sis­tency may be ex­plained by the fol­low­ing FAO claim:

“Es­ti­mates have been made for non-re­port­ing coun­tries as well as for coun­tries re­port­ing in­com­plete data. How­ever, in cer­tain coun­tries, data for chick­ens, ducks and turkeys do not yet seem to rep­re­sent the to­tal num­ber of these birds. Cer­tain other coun­tries give a sin­gle figure for all poul­try; data for these coun­tries are shown un­der “Chick­ens”.

It’s pos­si­ble that the FAO es­ti­mate for egg-lay­ing hens in­cludes egg-lay­ing birds other than chick­ens, while the es­ti­mate for chick­ens al­ive at any time does not. Another pos­si­bil­ity is that ei­ther Bangladesh offi­cial statis­tics un­der­es­ti­mate the num­ber of chick­ens in the coun­try, or FAO over­es­ti­mate the num­ber of egg-lay­ing hens.

Other coun­tries that ac­cord­ing to FAO data have more egg-lay­ing hens than chick­ens in gen­eral are North Korea, French Guiana, Latvia, and Ice­land. For Latvia and Ice­land both num­bers come from the offi­cial data. For North Korea and French Guiana the num­ber of chick­ens come from offi­cial data, and the num­bers of hens are FAO es­ti­mates. I re­ported these in­con­sis­ten­cies for all these coun­tries to the FAO and they promised to cor­rect it by the end of 2019.

The U.S. and the UK

FAO statis­tics for the U.K. seem to in­clude not only lay­ers, but also pul­lets (hens who are too young to lay eggs) and roost­ers (who are used for breed­ing hens).[4] FAO statis­tics for the U.S. in­clude lay­ers, but not pul­lets or roost­ers.[5] Pul­lets and roost­ers make up about 27% of all chick­ens in­volved in the U.S. egg in­dus­try[6] so the differ­ence is non-neg­ligible. It could be that FAO sim­ply does not spec­ify whether hen num­bers should in­clude pul­lets and roost­ers. Hence, differ­ent coun­tries re­port statis­tics in differ­ent ways which makes com­par­i­sons prob­le­matic. I re­ported this is­sue to the FAO and they promised to re­vise the data by the end of 2019.

Broilers (meat chicken)

Indonesia

Ac­cord­ing to FAO, in In­done­sia, in 2017 there were 2.18B chick­ens al­ive at any time (offi­cial data), 167M chick­ens in the egg-lay­ing hen in­dus­try (offi­cial data), and 2.85B chick­ens slaugh­tered for meat (“FAO data based on im­pu­ta­tion method­ol­ogy”). It would fol­low that the av­er­age slaugh­ter age of meat chicken is about 365 × (2.18B-167M) /​ 2.85B = 257 days,[7] which is much higher than any figure for broiler lifes­pan I’ve seen. Ac­cord­ing to Wright and Dar­mawan (2017), broilers in In­done­sia rarely live be­yond 35 days. Statis­ti­cal Year­book of In­done­sia 2017 (page 280) seems to claim that there are 1.6B broilers al­ive at any time, which is lower than the figure I es­ti­mated from FAOSTAT (2.18B-167M = 2B) but still much higher than we should ex­pect by look­ing at the slaugh­ter to­tal. I re­ported this is­sue to the FAO and they promised to re­vise the data by the end of 2019.

U.S.

Ac­cord­ing to FAOSTAT, there were 1.97B chick­ens al­ive at any time in the U.S. in 2017. Ear­lier in this ar­ti­cle I ex­plained that 505.8M of these chick­ens were used in the U.S. egg-lay­ing in­dus­try. It would fol­low that 1.97B − 505.8M = 1.47B broilers are al­ive at any time. Similarly, table 30 of the 2017 U.S. Agri­cul­ture cen­sus claims that there are 1.62B broilers al­ive at any time. FAOSTAT and USDA stats also claim that there were 9B chick­ens slaugh­tered in the U.S. in 2017. It would then fol­low that the av­er­age slaugh­ter age of broilers is 365 days × 1.47B /​ 9B = 59.6 days. How­ever, ac­cord­ing to the Na­tional Chicken Coun­cil (2019), the av­er­age broiler slaugh­ter age in the U.S. is 47 days. Ac­cord­ing to the same source, pres­laugh­ter mor­tal­ity of broilers is 5% which means that it can’t ex­plain the 24% differ­ence in slaugh­ter age es­ti­mates. Broiler moth­ers who live longer are already ex­cluded from the es­ti­ma­tion, so this can’t ex­plain it ei­ther. Live ex­ports and im­ports of chick­ens are also not big enough to ex­plain the differ­ence (69M live chick­ens im­ported, and 5M live chick­ens ex­ported ac­cord­ing to FAO).

How­ever, there is one pos­si­ble ex­pla­na­tion I’m aware of that could ex­plain the dis­crep­ancy. Between pe­ri­ods of rais­ing broilers, farms have down­time pe­ri­ods. Th­ese last about two weeks, dur­ing which fa­cil­ities are cleaned and pre­pared. Farms that are in a down­time when USDA is col­lect­ing data could be re­port­ing the num­ber of broilers that are in farm dur­ing the broiler rais­ing pe­riod in­stead of re­port­ing that the farm has no broilers. This would in­flate the num­ber of broilers who are al­ive at any time, but not slaugh­ter to­tals. How­ever, I’m far from cer­tain that this ex­pla­na­tion is cor­rect.

High var­i­ance in es­ti­mated slaugh­ter age

In this sec­tion, I will show that there is a very high var­i­ance in es­ti­mated slaugh­ter age of meat an­i­mals in var­i­ous coun­tries. This makes me sus­pect that there are similar in­con­sis­ten­cies in FAO’s data for some other coun­tries and an­i­mals.

I made a spread­sheet in which I very roughly es­ti­mated the slaugh­ter age of meat chick­ens, turkeys and pigs[8] in most of the coun­tries. I as­sumed that all pigs and turkeys were raised pri­mar­ily for their meat. I es­ti­mated their mean slaugh­ter age in each coun­try by di­vid­ing the num­ber of an­i­mals al­ive in 2017 by the num­ber of an­i­mals slaugh­tered in 2017. For chick­ens, the es­ti­ma­tion was very similar ex­cept that I also had to es­ti­mate the num­ber of meat chick­ens al­ive by sub­tract­ing the num­ber of egg-lay­ing hens from all the chick­ens al­ive in each coun­try. All the num­bers were taken from FAO.

Such an es­ti­ma­tion gives a very im­pre­cise mea­sure of slaugh­ter age for mul­ti­ple rea­sons:

  • An­i­mals are some­times raised in one coun­try and slaugh­tered in an­other one.[9]

  • It ig­nores pre-slaugh­ter mor­tal­ity.

  • If the num­ber of an­i­mals farmed in a coun­try is grow­ing or shrink­ing in a given year, it may dis­tort the es­ti­mate.

  • As ex­plained be­fore, egg-lay­ing hen num­bers for many coun­tries don’t cap­ture all the chick­ens used in egg-pro­duc­tion. Hence, if we sub­tract FAO’s num­ber for egg-lay­ing hens al­ive from FAO’s num­ber for all chick­ens al­ive in a given coun­try, we may over­es­ti­mate the num­ber of broilers.

  • Chicken slaugh­ter statis­tics for at least some coun­tries in­clude slaugh­ters of egg-lay­ing hens (see the Ap­pendix), which should ideally be ex­cluded for the pur­poses of this es­ti­ma­tion.

Nev­er­the­less, if all coun­tries sub­mit­ted data to FAO in a con­sis­tent man­ner, I’d ex­pect this es­ti­mated ra­tio of meat an­i­mals al­ive in 2017 to slaugh­tered in 2017 to be roughly similar in most coun­tries. How­ever, as it can be seen in Table 1, there is a very high var­i­ance in my es­ti­mated slaugh­ter ages in var­i­ous coun­tries.

Table 1: Roughly es­ti­mated mean slaugh­ter ages for se­lected coun­tries Table1

First 15 rows in the table are the biggest coun­tries by chicken slaugh­ters. The re­main­ing coun­tries are se­lected be­cause of some par­tic­u­larly sus­pi­cious val­ues.

The same species of an­i­mal is un­doubt­edly slaugh­tered at differ­ent ages in differ­ent coun­tries but I think the differ­ences are not nearly as dras­tic as the table sug­gests. Fur­ther­more, I see some non­sen­si­cal val­ues in the table. In some cases, they can be ex­plained by live im­ports and ex­ports[10] or a sud­den change in the in­dus­try size.[11] How­ever, in at least some cases they seem to in­di­cate in­con­sis­ten­cies in FAO data.[12]

Conclusions

The only ac­tion­able ad­vice here is that if you are bas­ing im­por­tant de­ci­sions (e.g. the coun­tries in which in which coun­tries to ex­pand your char­ity) on FAO num­bers, it may be worth ver­ify­ing whether or not those num­bers are cor­rect. For ex­am­ple, it could be use­ful to check my spread­sheet to see if slaugh­ter age es­ti­mates for the coun­try of in­ter­est seem to be sus­pi­ciously high or low. If they are, it may be worth look­ing deeper into the statis­tics or ask­ing about it on the FAOSTAT feed­back fo­rum.[13]

Ap­pendix: caveats about FAO’s an­i­mal statistics

Some egg-lay­ing hens are in the meat industry

FAO egg-lay­ing hen statis­tics for the U.K., the U.S., and prob­a­bly other coun­tries, in­clude broiler moth­ers. They are a part of the meat in­dus­try rather than the egg in­dus­try be­cause they lay eggs that hatch into chick­ens, who are raised for meat. Ac­cord­ing to my ex­trap­o­la­tions, there are 350M-600M broiler moth­ers in the world.[14]

Not all slaugh­tered chick­ens were raised for meat

Ac­cord­ing to FAOSTAT, there were 9,050M chick­ens slaugh­tered in the U.S. in 2017. Look­ing at the USDA statis­tics, it can be seen[15] that 135M of these slaugh­ters were of ma­ture chick­ens, who are not broilers. Most of them are “spent” egg-lay­ing hens who are slaugh­tered when their rate of egg-lay­ing slows down. Others are roost­ers used for chicken breed­ing. I’m un­sure if “spent” hens and roost­ers are in­cluded in FAOSTAT statis­tics for other coun­tries.

In the egg in­dus­try, male chicks are kil­led at a very young age (e.g. one day) be­cause they can not lay eggs, and are less suit­able for meat pro­duc­tion than broiler breeds. It seems that they are ex­cluded from the FAOSTAT chicken slaugh­ter to­tals.

An­i­mals slaugh­tered within the dairy industry

In the dairy in­dus­try, most male calves are slaugh­tered at a young age be­cause they can’t pro­duce milk, and older fe­male cows are slaugh­tered when they no longer pro­duce enough milk. By com­par­ing FAO and offi­cial U.S. statis­tics, I can see that slaugh­ters of both of these groups of an­i­mals are in­cluded in the FAO’s slaugh­ter to­tals for the U.S.[16] I haven’t checked the num­bers for other coun­tries and other species used for milk (sheep, goats, buf­faloes, camels, etc.).

References

Brienen, M., Cave­nagh, B., Van Vliet, W., Copier, M. (2014). Meet­ing the challenge of In­done­sia’s grow­ing de­mand for poultry

FAO Fish­eries Depart­ment (2001). Fish­ery Statis­tics: Reli­a­bil­ity and Policy Implications

Mass, B. L., Me­tre, T. K., Tsongo, F., Mugisho, A. B., Kam­pemba, F. M., Aya­girwe, R. B. B., Az­ine, P. C., & Chiuri, W. L. (2014). From taboo to com­mod­ity: his­tory and cur­rent situ­a­tion of cavy cul­ture in the Demo­cratic Repub­lic of the Congo. Live­stock Re­search for Ru­ral Devel­op­ment, 26, 151.

Mass, B. (2019). Africa: Why More Peo­ple in Africa Should Farm Guinea Pigs for Food

Nathan As­so­ci­ates Inc. (2013). In­done­sia’s Poul­try Value: Chain Costs, Mar­gins, Prices, and Other Issues

Na­tional Chicken Coun­cil. (2019). U.S. Broiler Perfor­mance.

Pauly, D., & Zel­ler, D. (2016). Catch re­con­struc­tions re­veal that global marine fish­eries catches are higher than re­ported and de­clin­ing. Na­ture com­mu­ni­ca­tions, 7, 10244.

Wald­horn D. (Forth­com­ing). Snails used for hu­man con­sump­tion. The case of meat and slime

Wright, T., Dar­mawan, B. (2017). In­done­sia Vol­un­tary Poul­try Re­port. USDA For­eign Agri­cul­ture Ser­vice, Global Agri­cul­tural In­for­ma­tion Net­work.

XIAO, H. B., Qiong, C. H. E. N., WANG, J. M., & Oxley, L. (2015). The puz­zle of the miss­ing meat: Food away from home and China’s meat statis­tics. Jour­nal of in­te­gra­tive agri­cul­ture, 14(6), 1033-1044.

Yu, X., & Abler, D. (2014). Where have all the pigs gone? In­con­sis­ten­cies in pork statis­tics in China. China Eco­nomic Re­view, 30, 469-484.

This es­say is a pro­ject of Re­think Pri­ori­ties. It was writ­ten by Saulius Šimčikas. Thanks to Cash Cal­laghan, David Moss, Ja­son Schukraft, Mar­cus A. Davis, Per­sis Eskan­der, and Sab­rina Ahmed for re­view­ing drafts of this post and mak­ing valuable com­ments.

Rethink Priorities Logo

Notes


  1. There is a longer ex­pla­na­tion about how data is col­lected in the method­ol­ogy doc­u­ment:

    In gen­eral, figures have been sup­plied by gov­ern­ments throught na­tional pub­li­ca­tions and FAO ques­tion­naires (both pa­per or elec­tronic). To make the cov­er­age of this data col­lec­tion as com­plete as pos­si­ble, offi­cial data have some­times been sup­ple­mented with data from un­offi­cial sources. Use has also been made of in­for­ma­tion sup­plied by other na­tional or in­ter­na­tional agen­cies or or­ga­ni­za­tions.

    ↩︎
  2. Maass (2019) claims that in Africa ro­dents are of­ten kept in houses or kitchens and are not in­cluded in most na­tional statis­tics. Fur­ther­more, the scale of ro­dent farm­ing in Africa could be rel­a­tively small. This makes their ex­clu­sion from FAO statis­tics un­der­stand­able. How­ever, it still fol­lows that FAO statis­tics provide an in­com­plete pic­ture on the topic of ro­dent farm­ing. ↩︎

  3. I see that hens are in­cluded in chicken num­bers for at least some coun­tries by com­par­ing FAO’s live an­i­mal num­bers for chick­ens in the UK gov­ern­ment web­site, Live­stock num­bers in UK spread­sheet, Poul­try tab. For all years since 2006 (ex­cept 2009), we can get the FAO num­ber for chick­ens in the U.K. by adding Hens and pul­lets lay­ing eggs for eat­ing, Breed­ing flock, and Table chick­ens (broilers) num­bers from the UK gov­ern­ment spread­sheet. For ex­am­ple, 158,202 thou­sand for 2006. For years af­ter 2009, FAO num­bers are rounded to the near­est mil­lion. The year 2009 seems to be in­putted in­cor­rectly and FAO’s num­ber for chick­ens seems to in­clude not only chick­ens, but also other poul­try (ducks, geese, turkey, etc.). ↩︎

  4. The num­bers in the UK gov­ern­ment web­site, Live­stock num­bers in UK spread­sheet, Poul­try tab, To­tal lay­ing and breed­ing fowl row seem to be very similar to num­bers in FAO’s live­stock pri­mary (“United King­dom”, “Pro­duc­ing An­i­mals/​Slaugh­tered”, “Eggs, hen, in shell”). For years un­til 2013 they are ex­actly the same. For ex­am­ple, 47,024,000 for 2013. From the UK gov­ern­ment’s spread­sheet it can be seen that these num­bers in­clude pul­lets and breed­ers. ↩︎

  5. The layer num­bers for the U.S. in the table “An­nual Aver­age Num­ber of Lay­ers, Eggs per Layer, and To­tal Egg Pro­duc­tion” from USDA’s Chick­ens and Eggs 2018 Sum­mary are iden­ti­cal to the ones in FAO’s live­stock pri­mary (“United States of Amer­ica”, “Pro­duc­ing An­i­mals/​Slaugh­tered”, “Eggs, hen, in shell”). For ex­am­ple, 375,845,000 for 2017. Ac­cord­ing to my cor­re­spon­dence with USDA “Any layer num­ber in the sum­mary re­port does not in­clude pul­lets. NASS’s defi­ni­tions of the two are mu­tu­ally ex­clu­sive: lay­ers are fe­males of any age lay­ing mar­ketable eggs while pul­lets are fe­males who are not yet lay­ing mar­ketable eggs.” Since the table in the USDA sum­mary says “lay­ers”, we know that it and FAO’s num­bers ex­clude pul­lets. ↩︎

  6. Ac­cord­ing to the table 30 in the Cen­sus of Agri­cul­ture, in 2017 in the U.S. there were 368.2M lay­ers, 130.5M pul­lets, and 7M roost­ers. So the to­tal num­ber of chick­ens in­volved in egg pro­duc­tion in the U.S. is about 505.8M. (130.5M + 7M) /​ 505.8M = 27.2%. ↩︎

  7. This might be a slight over­es­ti­mate be­cause it ig­nores pre-slaugh­ter mor­tal­ity. Some chick­ens are al­ive for some time but are not slaugh­tered be­cause they die of dis­ease or other causes be­fore their slaugh­ter age. How­ever, the broiler pre-slaugh­ter mor­tal­ity rate in In­done­sia is not nearly high enough to ex­plain these dis­crep­an­cies. Ac­cord­ing to Nathan As­so­ci­ates Inc. (2013), it’s 6-7 per­cent.” A similar rate can be in­ferred from Brienen et al. (2014). Ac­cord­ing to FAO, in 2017, In­done­sia im­ported 5.3M live chick­ens and ex­ported 7,000. If this was taken into ac­count, the es­ti­mated slaugh­ter age would in­crease even more (al­though very slightly). I didn’t take this into ac­count for the sake of sim­plic­ity. Other pos­si­ble prob­lems with this es­ti­mate are ex­plained in the High var­i­ance in es­ti­mated slaugh­ter age sec­tion. ↩︎

  8. I didn’t in­clude sheep in the spread­sheet be­cause sheep raised pri­mar­ily for wool are prob­a­bly slaugh­tered when they are older than sheep raised pri­mar­ily for meat. Con­se­quently, there might be big differ­ences in sheep slaugh­ter age in differ­ent coun­tries be­cause it de­pends on whether sheep in that coun­try are mostly raised for meat or wool. Un­for­tu­nately, FAO doesn’t col­lect statis­tics for how many an­i­mals are raised for wool since 2013. Similarly, I ex­cluded cat­tle from the table be­cause in many coun­tries they are used as work­ing an­i­mals and FAO does not col­lect statis­tics about it. ↩︎

  9. Stats about live im­ports and ex­ports are also pro­vided by FAO. In to­tal, about 1.8B chick­ens, 45M pigs, and 86M turkeys were trans­ported live be­tween coun­tries globally in 2017. Hence, it may cre­ate sig­nifi­cant dis­tor­tions. How­ever, I see no easy way to in­cor­po­rate them in the es­ti­ma­tions be­cause an­i­mals are trans­ported live to other coun­tries not just for slaugh­ter, but also for breed­ing and pro­duc­tion. More de­tailed statis­tics would be needed to do it (e.g. like the data for the EU). ↩︎

  10. E.g., ac­cord­ing to Table 1, meat chick­ens are slaugh­tered when they are just 7 days old in Sin­ga­pore, an im­plau­si­ble figure. Look­ing closer, in 2017 Sin­ga­pore had 3.6M chick­ens al­ive at any time, of which 2.8M were egg-lay­ing hens, slaugh­tered 46.8M chick­ens, and im­ported 46.8M chick­ens al­ive. The most plau­si­ble ex­pla­na­tion is that chick­ens are im­ported to Sin­ga­pore just to be slaugh­tered, hence there is no in­con­sis­tency. ↩︎

  11. E.g., ac­cord­ing to my es­ti­ma­tion, pigs in Swe­den are slaugh­tered when they are just 20 days old. This is be­cause ac­cord­ing to FAO statis­tics, in 2017 Swe­den slaugh­tered 2.6M pigs but had only 138K pigs al­ive at any time on av­er­age. How­ever, in 2016 Swe­den had 1.5M pigs al­ive at any time on av­er­age. If many of these pigs were slaugh­tered early in 2017 and not re­placed with new pigs, it could at least partly ex­plain the dis­crep­ancy (live im­ports and ex­ports of pigs were very low in both years). ↩︎

  12. E.g., ac­cord­ing to Table 1, meat chick­ens are slaugh­tered in Ethiopia when they are al­most 3 years old. Ac­cord­ing to FAO, in 2017 in Ethiopia, 17.1M chick­ens were slaugh­tered for meat and there were 59.5M meat chick­ens al­ive at any time of which 11M were egg-lay­ing hens, 0.4M chick­ens were im­ported live and there is no en­try for chicken ex­ports. Num­bers for pre­vi­ous years are quite similar so that can­not ex­plain the in­con­sis­tency. Hence, it could be that at least one of these statis­tics is in­ac­cu­rate. Another ex­am­ple: ac­cord­ing to Table 1, turkeys in Den­mark are slaugh­tered when they are more than 300 years old. Look­ing closer, ac­cord­ing to FAO, in 2017 there were 304,000 turkeys in Den­mark at any time, 1,000 turkeys slaugh­tered, 2.7M live turkeys im­ported and 872,000 live turkeys ex­ported. I can’t make sense of these num­bers, hence it’s pos­si­ble that at least one of them is in­ac­cu­rate. ↩︎

  13. Note that when you ask on the FAOSTAT feed­back fo­rum, ques­tions need to be ap­proved which can take weeks. Fur­ther­more, all ques­tions in the fo­rum are reg­u­larly deleted for some rea­son. Hence, af­ter post­ing a ques­tion it is ad­vis­able to check if it’s an­swered pe­ri­od­i­cally. ↩︎

  14. I haven’t found a de­pend­able es­ti­mate of how many broiler moth­ers there are in the world. I tried to ex­trap­o­late based on statis­tics for var­i­ous years and re­gions (U.S., UK, EU, South Africa, Brazil, Pak­istan). Ex­trap­o­la­tions as­sumed that the ra­tio be­tween chicken slaugh­ters and broiler breed­ers al­ive at any time is similar in differ­ent coun­tries and years. I’m un­sure to what de­gree this as­sump­tion is cor­rect. If broiler moth­ers in differ­ent coun­tries lay a similar amount of eggs per month, these es­ti­ma­tions should be roughly cor­rect. They might still be a bit in­ac­cu­rate be­cause:

    • I don’t take broiler mor­tal­ity rates into ac­count and they differ by country

    • Not all chicken slaugh­ters that are in­cluded in the FAO statis­tics are broilers. E.g., U.S. statis­tics in­clude “spent” hens, who are slaugh­tered when their egg pro­duc­tion slows down. For some coun­tries, chicken slaugh­ter to­tals also in­clude other slaugh­ters of other species of birds.

    ↩︎
  15. U.S. statis­tics seem to in­clude some of the slaugh­tered spent hens and none of cul­led day-old male chicks. USDA’s Poul­try Slaugh­ter re­port and FAO re­port the ex­act same num­ber of chick­ens slaugh­tered in the U.S. in 2017: 9,050,716,000. Ac­cord­ing to USDA, 8,916,097,000 of these chicken were “young”. “Young” is defined as “Com­mer­cially grown broilers-fry­ers and other young im­ma­ture birds such as roast­ers and capons.” Since cul­led male chicks are not men­tioned, I as­sume that they are ex­cluded. “Ma­ture” chick­ens are defined as “fowl from breeder and mar­ket egg flocks and stags and cocks”. This seems to in­clude spent hens from the egg-lay­ing in­dus­try. How­ever, ac­cord­ing to the USDA’s re­port, there were 134,619,000 ma­ture chick­ens slaugh­tered in 2017. I would have ex­pected that there would be a higher num­ber of egg-lay­ing hens slaugh­tered be­cause there were 375,845,000 egg-lay­ing hens in the U.S. on av­er­age and they live 1.5 to 2 years. This makes me think that not all spent hens were in­cluded. Note that all of this is about the U.S., and other coun­tries could be re­port­ing the data differ­ently. ↩︎

  16. See the table “Com­mer­cial and Farm Slaugh­ter by Species – United States: 2017 and 2018” in Live­stock Slaugh­ter 2018 Sum­mary. If we add 32,280.6 thou­sand cat­tle slaugh­tered in 2017 and 536.8 calves, we get that 32,817.4 thou­sand cat­tle were slaugh­tered in 2017 in the U.S. in to­tal. That is the ex­act figure we see in FAO’s Live­stock Pri­mary statis­tics if we choose “United States of Amer­ica”, “Pro­duc­ing An­i­mals/​Slaugh­tered”, “Meat, Cat­tle”, and “2017”. From the table Fed­er­ally In­spected Slaugh­ter and Per­cent by Clas­sifi­ca­tion and Month – United States: 2018 and 2017 To­tal of Live­stock Slaugh­ter 2018 Sum­mary we can see that dairy cows are in­cluded in the statis­tics. ↩︎