Prepared for Frontline Farming by Evan Gallant
Introduction
Colorado’s population is rapidly growing, having increased by 14.8% between 2010 and 2020 (Bradbury, 2021) -- a higher rate than the US’s 12.2% in the same timeframe. This rate is the 5th highest rate in the nation, with almost all growth occurring along the front range’s urban corridor (Fish, Najmabadi, and Vo, 2021). As a largely agricultural state with approximately half of the total land area dedicated to various forms of agriculture, we were curious how representative the industry has been (historically and currently) of the state’s BIPOC population (Farmland Information Center, 2021). Originally our goal was to find data describing the total number of BIPOC farmland owners at the county level - unfortunately, this data is either not available through the Census of Agriculture or was never collected. The most relevant extant data describes the number of farms on which BIPOC producers have worked, and the total land area covered by those farms. The distinction here is that “producers” do not necessarily own the farmland they work on. So, while this data does provide insight into the representation of BIPOC individuals making managerial decisions on the state’s farmland, it importantly does not discuss ownership nor representation of other BIPOC involved in food production such as farm or ranch workers. We provide this data set and analysis as a foundation from which more detailed analyses can be conducted, looking either at specific counties or specific BIPOC groups in Colorado.
Definitions
The Census of Agriculture has defined various BIPOC groups inconsistently over the last 50 years. Since racial classifications are fundamentally arbitrary and socially defined, tracking collective equity becomes a challenge as these categories are reported differently over time. While racial classifications are anchored to legal decisions including “OZAWA V. UNITED STATES (1922)” (Cornell Law School, 2021) and “THIND V. UNITED STATES (1923)”, these too are subject to shifts coinciding with fluctuations in social identities. A marked shift occurred in 2007, when multiple groups were reclassified and renamed. White and Black producers were defined consistently in the Census; definitions for the remaining groups as they are categorized in the research are listed below.
Group Name | Definition |
American Indian/ Alaskan Native | This group is an official category starting in 2007, but only includes “American Indians” in the 1978, 1987, and 1997 censuses. It is unknown if Alaskan Native individuals were reported in a different group before 2007, or if they were referred to as American Indian peoples as well. |
Asian/ Pacific Islander | This group is an official category in the 1978, 1987, and 1997 Censuses. For the purposes of this analysis, "Asian Operators" and "Native Hawaiian or Other Pacific Islander Operators" from the 2007 and 2017 censuses were combined into one group. |
Hispanic, Latino, and Spanish Origin | This group is an official category starting in the 2007 Census, and a combination of the two groups "All Other Races" and "Spanish Origin" in the 1978, 1987, and 1997 Censuses. (See “All Other Races” below) |
All Other Races | From the Census: “This category is primarily limited to persons native to or of ancestry from Mexico, the Caribbean, and Central and South America”, and was in use prior to 2007. |
Additional definitions relevant to understanding the data set are listed below:
Term | Definition |
Producer/ Operator | The term "operator" (before 2017) or “producer” (starting in 2017) designates a person who operates a farm, either doing the work or making day-to-day decisions about such things as planting, harvesting, feeding, and marketing. The operator/producer may be the owner, a member of the owner’s household, a hired manager, a tenant, a renter, or a sharecropper. If a person rents land to others or has land worked on shares by others, he/she is considered the operator/producer only of the land which is retained for his/her own operation. |
Land in Farms | Land in farms is an operating unit concept and includes land owned and operated as well as land rented from others. |
Methods
All data were collected or calculated from the Colorado Census of Agriculture for the years 1978, 1987, 1997, 2007, and 2017. Specific data sets with links are listed below:
A notable caveat of the data exists due to the number of producers that may be present on a single farm: up to 4 or 5 producers (depending on the year) may be counted in the census for each farm, meaning that if multiple individuals from different BIPOC groups work together on a single farm, that farm will show up multiple times in calculations which require the totalling of all BIPOC groups’ data. This was significant in determining the values for the “White Number of Farms” and “White Land in Farms” between 1978 and 1997. These two categories were not explicitly reported in the 1978, 1987, or 1997 censuses, and were instead calculated by summing the Number of Farms and Land in Farms values (respectively) from all BIPOC groups for each year, then subtracting that total from the overall state totals for Number of Farms and Land in Farms:
(Spanish et al. # of Farms + American Indian/Alaskan Native # of Farms + Asian # of Farms + Black # of Farms + Hawaiian/Pacific Islander # of Farms) - Colorado # of Farms = White # of Farms
(Spanish et al. Land in Farms + American Indian/Alaskan Native Land in Farms + Asian Land in Farms + Black Land in Farms + Hawaiian/Pacific Islander Land in Farms) - Colorado Land in Farms = White Land in Farms
Due to the potential unknown overlap of producers from multiple non-white groups being reported on a single farm, these calculations should not be used for statistical analysis.
The (D) Hurdle
In the Census of Agriculture some data are suppressed and shown as “(D)”, meaning these values are “withheld to avoid disclosing data for individual operations” (nass.usda.gov, 2019). This most commonly occurs when there are a small number of farms and/or producers within a county, where reporting specific values could lead to the identification of those producers. For instance, in 2017 three farms with Black producers existed in Boulder county, but (D) was reported for the Land in Farms value in that year:
County | 2017 Number of Farms | 2017 Number of Producers | 2017 Number of Principal Producers | 2017 Land in Farms |
Boulder | 3 | 3 | 3 | (D) |
While this obfuscation is necessary for anonymity, it prevented consistent analysis on the county scale, especially for areas without many BIPOC producers.
Discussion
Significant trends at the state level, by racial group, and for specific counties are discussed below. These analyses are based on comparisons between the Number of Farms and Land in Farms census categories for each racial group between 1978, 1987, 1997, 2007, and 2017.
State of Colorado Trends
While all racial groups reported in the Census of Agriculture (including White, Black, Hispanic Latino and Spanish Origin, American Indian/Alaskan Native, and Asian/Pacific Islander) experienced growth in the number of farms on which a producer of their group worked (Number of Farms), only the American Indian/Alaskan Native and Asian/Pacific Islander groups experienced growth in the acres of land on which a producer of their group worked (Land in Farms). This indicates either a trend of decreasing farm size despite increases in the total number of producers working on those farms, or a concentration of farm labor on existing large farms combined with the foreclosure/selling/retirement of smaller farms.
BIPOC representation on farms has increased from 1978 to 2017: in 1978 4.32% of farms had BIPOC producers on them, and 5.93% of the total farmland was operated at least in part by BIPOC operators. In 2017 these rates rose to 10.29% and 7.79%, respectively.
Racial Group Trends
White - Largest Producer Group (by Number of Farms): While the number of farms with White producers increased by 9,973 between 1978 and 2017, the total land farmed by White producers decreased by 2,511,287 acres. Acreage peaked in 1978, while the number of farms peaked in 2017. White producers far outnumbered producers from any of the BIPOC groups, composing the majority of the agricultural workforce in Colorado.
Hispanic, Latino, and Spanish Origin - 2nd Largest Producer Group (by Number of Farms): Hispanic, Latino, and Spanish origin producers followed a similar trend to Black producers, although with significantly higher numbers throughout analyzed years. In 1978 this group represented 973 farms on 1,662,208 acres of land. After a sharp decline in 1987 (see the discussion on Montezuma county below for more information), numbers slowly increased in the following decades, reaching 3,050 farms on 1,302,256 acres of land in 2017. Similarly to the Black producer group, while the number of farms with Hispanic, Latino, and Spanish origin producers increased (by 2077), the total acreage for the group decreased over the same time period (by 359,952 acres).
American Indian/Alaskan Native - 3rd Largest Producer Group (by Number of Farms): The American Indian/Alaskan Native group was one of the two racial groups which experienced increases in both the number of farms (361 more), and total acreage farmed (600,231 acres more) over the analyzed time period.
Asian/Pacific Islander - 4th Largest Producer Group (by Number of Farms): Asian/Pacific Islanders were the other group who experienced increases in both the number of farms (195 more), and total acreage farmed (51,531 acres more) over the time period. Although the total amount of land with Asian/Pacific Islander producers peaked in 2007 at 136,148 acres, the number of farms did not follow the same trend, instead peaking in 2017 while Land in Farms decreased.
Black - Smallest Producer Group (by Number of Farms): Black producers have consistently represented a small portion of the state's agricultural workers. While the number of farms with Black producers on them increased by 30, from 56 to 86, between 1978 and 2017, the total land farmed by Black producers decreased 3,296 acres, from 14,035 to 10,739. After decreasing to a low of 19 farms and 4,664 acres farmed in 1987 after an initial peak in 1978, numbers slowly increased in the following decades until the present, although the total acreage farmed never reached initial 1978 values.
County trends
Montezuma - Hispanic, Latino, and Spanish Origin: In 1978 565,167 acres were reported for the "All Other Races" category (persons native to or of ancestry from Mexico, the Caribbean, and Central and South America), and 566,043 acres for the "Spanish Origin" category. Even if all acres reported were double-counted for the two groups (i.e. each farm and its corresponding acreage had at least 1 producer from each group), a precipitous decline occurred between 1978 and 1987. In 1987 2,216 acres and 2,705 acres were reported respectively for these groups.
Montezuma - American Indian/Alaskan Native: While no data was available for the American Indian/Alaskan Native group in 1978 for Montezuma county, in 1987 they represented 8 farms on 593,638 acres. The total acreage for "All Other Counties", where Montezuma's data would reside for 1978, is 41,350, meaning the county's acreage would have to have increased at least 552,288 acres over that decade. This single county accounted for more than half of the group's acreage total for 1978 in the state. While the acreage for farms in Montezuma county is hidden in the later data sets (listed as (D) for personal confidentiality reasons), the acreage represented by (D) listings in Montezuma county for 2017 totals 669,313 acres, which again is more than half of the state total for that decade. This trend is notable due to the opposite occuring with the Hispanic, Latino, and Spanish origin group over the same time period.
La Plata - American Indian/Alaskan Native: In 1978 and 1987 this group's Land in Farms value increased from 295,687 to 297,360 acres, while their Number of Farms decreased from 22 to 6. In the following years, while the Land in Farms values were hidden as (D), the Number of Farms increased again, peaking at 35 in 2007. This sharp decline and sudden increase in the Number of Farms is interesting and worthy of further investigation, given La Plata County’s inclusion of the Southern Ute Reservation.
State Producer Trends
The graph below displays the growth rate of the number of producers by race in a decade-over-decade comparison for the state of Colorado. This data was not disaggregated by county, and thus was analyzed separately from county-specific data on the Number of Farms and Land in Farms for each racial group.
Fig 1: Change in Producer Growth Rates by Race (Decade over Decade)
This graph highlights the significant growth for all groups of producers between 1997 and 2007. Despite differences in the raw number of producers from each group, their rates of change are similar between each decade. The outlying peak in the American Indian/Alaskan Native group’s growth rate between 1997 and 2007 may be correlated with the significant increase in the number of farms with American Indian/Alaskan Native producers working on them in La Plata County during that decade. More investigation into the history of land use in La Plata during that time period could clarify this extremely high growth rate.
Additionally, it is noteworthy that only Black (the smallest group) and White (the largest group) producers experienced growth during all decades examined. While the other groups experienced higher growth rates between 1997 and 2007, they also exhibited decreasing trends in either the 1987-1997 or 2007-2017 time periods.
Further Research to be Conducted
Two counties, Montezuma and La Plata, exhibit interesting trends for the Hispanic, Latino, or Spanish origin and American Indian/Alaskan Native groups. The following questions could help guide further research to identify what factors or events contributed to these trends, and illuminate the potential effects of similar changes in Colorado’s agriculture today.
County | Identity | Research Question | Context |
Montezuma | Hispanic, Latino, or Spanish Origin | What explains the 1,128,495 acre drop for the Hispanic, Lation, or Spanish Origin group in Montezuma between 1978 and 1987? | Hispanic, Latino, or Spanish Origin Land in Farms values drop precipitously, then never recover. Possible explanations could include a change in the census reporting structure, a change in self-described cultural identity within the county, loss or selling of large amounts of land, or an error in the census. |
Montezuma | American Indian/ Alaskan Native | Do Land in Farms values follow the same increasing trend as Number of Farms values for the American Indian/Alaskan Native group? | The only year with non-hidden Land in Farms data for the American Indian/Alaskan Native group is 1987. In this year the total acreage for the group in Montezuma is 593,638, with only 8 farms reported. In the following years, while acreage is hidden, the Number of Farms value increases significantly. Is there a way to find out if acreage increases proportionally? |
Montezuma | Black | What explains the 2,781 acre drop for Black producers between 2007 and 2017? | While Black producers are not represented in Montezuma's data until 2007, between 2007 and 2017 their Land in Farms value decreased from 2805 to 24. Interestingly, during this same time period their Number of Farms value increased from 3 to 6. |
La Plata | American Indian/ Alaskan Native | Do Land in Farms values follow the same increasing trend as Number of Farms values for the American Indian/Alaskan Native group after 1987? What is the relationship between Number of Farms and Land in Farms, given the non-linear relationship seen when 1978 and 1987 are compared? | The only years with non-hidden Land in Farms data for the American Indian/Alaskan Native group are 1978 and 1987. In these years the total acreages for the group in La Plata are 295,687 and 297,360 respectively, with 22 and then 6 farms reported. In the following years, while acreage is hidden, the Number of Farms value increases significantly, surpassing the 1978 value. Is there a way to find out if acreage increases proportionally, especially considering the significant drop in Number of Farms between 1978 and 1987 in comparison to the lack of change in the Land in Farms values between those two years? |
Additionally, the data analyzed in this project does not include ownership rates for BIPOC groups, as the only available relevant census categories which were reported on the county level described the number of farms and producers, and the acreage of those farms. While this data does highlight interesting trends, it does not allow us to discuss farmland ownership, which is more intrinsically tied to the representation of individuals within the agricultural decision making process. Additional research into other statistical sources which may include ownership data could lead to a more robust analysis of this topic.
References
“American Indian or Alaska Native Operators: 2007.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2007/Full_Report/Volume_1,_Chapter_2_County_Level/Colorado/st08_2_051_051.pdf.
“American Indian or Alaska Native Producers: 2017.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2017/Full_Report/Volume_1,_Chapter_2_County_Level/Colorado/st08_2_0049_0049.pdf.
“Asian Operators: 2007 & Black or African American Operators: 2007.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2007/Full_Report/Volume_1,_Chapter_2_County_Level/Colorado/st08_2_052_053.pdf.
“Asian Producers: 2017 & Black or African American Producers: 2017.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2017/Full_Report/Volume_1,_Chapter_2_County_Level/Colorado/st08_2_0050_0051.pdf.
Bradbury, S., & Burness, A. (2021, August 17). A growing, more diverse Colorado: 9 key takeaways from new census data. Retrieved from https://www.denverpost.com/2021/08/13/colorado-census-data-population-growth/
“Colorado Data and Statistics.” FIC, 2021, farmlandinfo.org/statistics/colorado-statistics/. Accessed 11 Oct. 2021.
“Here’s How Colorado Grew in the Last Decade, according to the Census.” The Colorado Sun, 13 Aug. 2021, coloradosun.com/2021/08/13/colorado-2020-census-population-growth/.
“Hispanic, Latino, or Spanish Origin Producers: 2017.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2017/Full_Report/Volume_1,_Chapter_2_County_Level/Colorado/st08_2_0048_0048.pdf.
“Native Hawaiian or Other Pacific Islander Operators: 2007.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2007/Full_Report/Volume_1,_Chapter_2_County_Level/Colorado/st08_2_054_054.pdf.
“Native Hawaiian or Other Pacific Islander Producers: 2017.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2017/Full_Report/Volume_1,_Chapter_2_County_Level/Colorado/st08_2_0052_0052.pdf.
“Operators by Selected Racial Groups: 1978.” National Agricultural Statistics Service, lib-usda-05.serverfarm.cornell.edu/usda/AgCensusImages/1978/01/06/182/Table-41.pdf.
“Operators by Selected Racial Groups: 1987 and 1982.” National Agricultural Statistics Service, lib-usda-05.serverfarm.cornell.edu/usda/AgCensusImages/1987/01/06/4/Table-34.pdf.
“Operators by Selected Racial Groups: 1997 and 1992.” National Agricultural Statistics Service, lib-usda-05.serverfarm.cornell.edu/usda/AgCensusImages/1997/01/06/1600/Table-37.pdf.
“Operators of Spanish, Hispanic, or Latino Origin: 1997 and 1992.” National Agricultural Statistics Service, lib-usda-05.serverfarm.cornell.edu/usda/AgCensusImages/1997/01/06/1600/Table-38.pdf.
“Selected Characteristics of Farms Operated by Females, Persons of Spanish Origin, and Specified Racial Groups: 1987 and 1982.” National Agricultural Statistics Service, lib-usda-05.serverfarm.cornell.edu/usda/AgCensusImages/1987/01/06/3/Table-17.pdf.
“Selected Characteristics of Farms by Specified Racial Groups, Sex of Operators, and Persons of Spanish, Hispanic, or Latino Origin: 1997 and 1992.” National Agricultural Statistics Service, lib-usda-05.serverfarm.cornell.edu/usda/AgCensusImages/1997/01/06/1599/Table-17.pdf.
“Selected Operator Characteristics by Race: 2007.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2007/Full_Report/Volume_1,_Chapter_1_State_Level/Colorado/st08_1_057_057.pdf.
“Selected Producer Characteristics by Race: 2017.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2017/Full_Report/Volume_1,_Chapter_1_State_Level/Colorado/st08_1_0063_0063.pdf.
“Spanish, Hispanic, or Latino Origin Operators: 2007.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2007/Full_Report/Volume_1,_Chapter_2_County_Level/Colorado/st08_2_050_050.pdf.
“TAKAO OZAWA v. UNITED STATES.” LII / Legal Information Institute, 2019, www.law.cornell.edu/supremecourt/text/260/178.
“UNITED STATES v. BHAGAT SINGH THIND.” LII / Legal Information Institute, 2019, www.law.cornell.edu/supremecourt/text/261/204.
“USDA - National Agricultural Statistics Service - Census of Agriculture - Frequently Asked Questions 2017.” Www.nass.usda.gov, www.nass.usda.gov/AgCensus/FAQ/2017/index.php.
“White Operators: 2007.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2007/Full_Report/Volume_1,_Chapter_2_County_Level/Colorado/st08_2_055_055.pdf.
“White Producers: 2017.” National Agricultural Statistics Service, www.nass.usda.gov/Publications/AgCensus/2017/Full_Report/Volume_1,_Chapter_2_County_Level/Colorado/st08_2_0053_0053.pdf.