CTA - Brazil Farmer Survey

Ryan Delany
Nov 7, 2023
5 min read

Section 1. Key Findings

Over the course of Oct 24th through 31st , 2023 we conducted a survey of Brazilian coffee farmers about their current crop, farms, farming practices and their expectations for the coming crop. We received 72 responses from farmers ranging from 1 hectare farms to 5,000 hectare farms across all of the major coffee growing regions in Brazil, with the exception of Paraná. The total hectarage covered in the survey represents roughly 1% of the total Brazil coffee hectarage in Brazil according to CONAB.

This extensive coverage provided some statistically significant findings that have altered our impressions of the 23/24 crop in particular and given us cause to adjust our production numbers and expectations for the 24/25 crop and beyond.

23/24 crop production is estimated to be 14.8% larger than the 22/23 crop (+17.57% for “my farm and + 12.1% for “my region”).
24/25 crop potential looks very positive, flowering was reported better than average by a majority of respondents 29 vs (26 reporting normal and 16 worse than average) and branch growth and plant health are both reported strong.
Farmers report good husbandry with 17 reporting using more fertilizer than normal vs only 6 who reported using less (49 report normal fertilizer use).

Section 2. Overview of Results

A. Production – 23/24 Crop

Farmer Estimates of Production 23/24 crop vs 22/23 crop the largest group was +50% or better that was 19 respondents or 26.38%.

Farmers were more optimistic of their own farms vs the region as a whole. The weighted average of the 23/24 crop was +17.57% for “My Farm” vs +12.1% for “My Region”. This could be selection bias where the type of farmers who fill our surveys and are connected to our networks tend to take better care of their farms, or perhaps more likely, farmers tend to be more optimistic about their own crops than others. Either way we could take an average of the two +14.8% and consider this to be a reasonable estimation of the size of the crop.

Fig. 2 - 23/24 Production Estimation

If we apply this +14.8% to our 22/23 crop number of 59.28 million, we would get a 23/24 crop number of 68.05 million, this is a bit more than our current estimate of 65.6 and we will be raising our estimate accordingly. This would raise our global Arabica surplus from a somewhat balance view of +0.73 Arabia and +3.89 total coffee to +2.31 Arabica Surplus and +6.34 total coffee surplus. This seems an appropriate increase based on the bearish price action of the last year.

Photo 1 – 23/24 Crop

B. Production – 24/25 Crop

For the 24/25 crop the data is less formal. We can talk about the flowering for the new crop in general terms and say that it is positive. Given 5 choices for flowering: Excellent, Above Normal, Normal, Below Normal, and Bad. The bulk of the answers were positive but not overwhelmingly so. 29 answers were positive vs 16 that were negative.

If we consider excellent to be 2 points, above normal to be 1 point, normal to be 0, below normal to be –1 and bad to be –2, then the weighted average was +0.26.

Fig. 3 – Flowering 24/25 Crop

Photo 2 – 23/24 Matas de Minas

Photo 3 – 23/24 Espírito Santo

We can speculate that the relatively high number of below average flowerings could be on account dryness that occurred early in the flowering cycle. Notably, below avg flowering was reported only in Minas Gerais (Arábica) and Espírito Santo (Robusta), specifically: Sul de Minas (9), Cerrado (3), Matas de Minas (3) and Espírito Santo (1). All the other regions reported flowering to be either normal or better than normal to excellent.

Farmers are making good use of high prices to increase husbandry on their farms with 17 reporting more fertilizer use than normal vs only 4 who reported using less. This further supports an impressive crop for 24/25. Using the same point scale as above the weighted average is +0.18.

Fig. 5 - Fertilizer Usage

Looking at the branch growth potential for the 24/25 crop also looks rather good. 29 of the answers were positive vs only 11 that were negative. The weighted average (using the same point scale as above) was 0.55, which is rather optimistic.

Fig. 4 - Branch Growth

Fig. 6 - Plant Health

Farmers who were pessimistic about new branch growth and/or plant health, are located in either Minas Gerais (9) or Bahia (2). We only had 3 Bahia respondents, which is a small sample to base conclusions, but the fact that 2 of them are rather pessimistic about the new growth is noteworthy.

Section 3. Scope and Limitations

A. Limitations:

In reviewing the data, we noticed that some of our questions could have been formulated better. For example, we asked about pruning, but our answer buckets were so large and our wording so vague that it is difficult to draw any serious conclusions from the information. In our next survey we will improve the wording to focus on pruning done over the past crop year and add more granularity to the potential responses.

Another major limitation is that our survey does not have many respondents from the Robusta-heavy region of Espirito Santo. Conab puts the production of the region at 23.3% but only 4% of our respondents reported being from ES. By contrast, Minas is over-represented in our survey as 83% of our respondents' reported farms in Sul De Minas, Matas de Minas or Cerrado vs 51% according to CONAB. This does imply that our data is heavily skewed towards the observations in those regions.

However, of the 3 farms that reported from ES, all 3 were very optimistic on 23/24 production and on flowering for the 24/25 crop, so it does support the overall optimistic view on production.’

B. Scope of the Survey:

Survey was conducted in Portuguese via Microsoft form over a 1-week period from 10/24/2023 - 10/31/2023. Farmer contact information was gathered through connections to farmers from personal connections to CTA as well as social media outreach.

We received 71 responses from 8 different regions with most respondents (32) reporting Sul de Minas and only 1 response from Rondonia.

Fig. 8 - Farm Location

In total hectares, the survey covered 17.5k hectares which is just under 1% (0.93%) of average production area reported by CONAB over the last 7 years.