Combating the production and distribution of fraudulent pesticides is a continuous concern for all actors of the pesticide supply chain in order to reduce their adverse impacts on the industry, farmers, and the environment. The present study assessed the current level of farmers’ perceptions, purchasing, and recognition regarding fraudulent pesticides. Furthermore, this study assessed how farmers with behavioral heterogeneity and different socioeconomic characteristics purchase fraudulent pesticides in the study area. The insights into farmers’ behaviors regarding fraudulent pesticides from this assessment enrich existing literature in the field, which are mainly based on measuring the quality of these pesticides as compared to authentic products. Additionally, the study provides a series of guidelines for designing extension programs and advisory services to raise awareness among farmers. While farmers’ actual purchasing rates of fraudulent pesticides have not previously been documented in Egypt, these results are consistent with the country’s 2030 vision of developing an active framework for tackling counterfeit and illegal pesticides on the agricultural market [46].

A growing number of purchases of fraudulent pesticides was observed in the study area. Our findings highlight that an overwhelming number of farmers have suffered from non-deceptive purchasing. The results are consistent with those of Kassem and Alotaibi [12], who found that 73.5% of farmers have purchased fraudulent pesticides in Saudi Arabia. In Uganda, a study conducted by Ashour et al. [17] found that 41% of farmers believe that local herbicides are counterfeit. This study also showed that among a large sample of herbicides collected, 30% contained less than 75% of the active ingredient advertised. Furthermore, another study conducted a market survey of fraudulent pesticides in Mali and found that counterfeit and illegal pesticides accounted for approximately 26% of all pesticide volumes sold [16]. Another market survey conducted in Egypt to monitor the counterfeit situation for a pesticide widely used in Egypt (abamectin) [39] showed that among the samples collected, 42.86% were not registered through the Egyptian Agricultural Pesticides Committee. Additionally, the percentage of the active ingredient in 71.4% of the samples was less than the acceptable limit, while 14.3% of the samples did not contain any abamectin. This result might be due to the fact that more than half of the farmers did not have good education (Table 1). Less-educated farmers do not have the capability of increasing their knowledge about the efficiency of counterfeits or reading label information, or, consequently, differentiating between original and fake products. Moreover, this result might also be because the majority of farmers rely on pesticide retailers as the main source of information about pesticides (Table 1). Pesticide dealers may not be considered as a trusted source for information in some cases. As indicated by Lekei et al. [47], some pesticide retailers, particularly unauthorized shops, may deceive farmers about the performance and effectiveness of pesticides in order to guide them toward certain products in pursuit of profit and as a kind of product advertisement. Meanwhile, as noticed during the field study, some dealers offer pesticides to farmers on a credit basis. This method may build trust-based relationships. The findings also highlighted that among those farmers who purchased fraudulent pesticides, the majority of them expressed that they had purchased such pesticides due to their low price. This might be a result of those farmers purchasing these pesticides for their price advantage in the belief that their quality is satisfactory. This result is consistent with the results of [28], who confirmed that the financial and accessibility criteria ranked second after the performance and effectiveness criteria in terms of importance when farmers select and use pesticides. Therefore, pesticide companies should adopt price-related anti-counterfeiting measures by implementing two mechanisms: first, by introducing lower-price product entry lines to reduce price gaps; second, by reviewing and reducing market, transaction, and production costs to minimize the risk of others undercutting the cost of the product [48,49,50].

Variable	Frequency (n = 468)	%
Age (min. = 30; max. = 79; mean = 56.23; SD = 9.04)
<45 years	33	7.1
45–60 years	296	63.2
>60 years	139	29.7
Education
Illiterate	163	34.8
Read and write	137	29.3
Basic education	36	7.7
Secondary school	52	11.1
University	80	17.1
Farm size (min. = 0.4; max. =8; mean = 2.61; SD = 1.73)
<1 hectare	261	55.8
1–3 hectares	106	22.6
>3 hectares	101	21.6
Off-farm income
Yes	266	56.8
No	202	43.2
Farming experience (min. = 5; max. = 55; mean = 25.72; SD = 10.51)
<16 years	101	21.6
16–30 years	239	51.1
>30 years	128	27.4
Type of agricultural production
Crops	225	48.1
Vegetables	80	17.1
Fruits	163	34.8
Information sources on pesticides *
Other farmers	197	42.1
Extension	183	39.1
Input dealers	329	70.3
TV and social media	133	28.4
Private companies	119	25.4
Pesticide’s label	97	20.7
Training experience in pesticides
Yes	82	17.5
No	386	82.5

However, the use of visual inspection for detecting fraudulent products is the first step in an overall strategy against fraudulent pesticides, even though this approach is not reliable for detecting counterfeits in most cases. In this context, our findings show that most farmers identify fraudulent pesticides based on their low efficiency, which may indicate that there is resistance to it. This means that farmers face difficulties in accurately identifying fraudulent pesticides during phase of the cycle from purchasing to spraying. This might be because farmers lack sufficient knowledge of the different methods of counterfeiting. Furthermore, we doubt that a farmer would go through a list of chemicals to check that everything was legal. According to Figure 2 in this paper, most farmers do not know how to clearly detect frauds through a visual assessment of the product label and packaging. Furthermore, according to Figure 1, over three-fourths of farmers make their assessment after spraying in their fields, based on the poor results obtained. But many other factors affect spraying efficacy—including time of day, wind speed, and nozzle settings. Poor application methods rather than fraud are possibly to blame for poor pest control results. Ashour et al. [17], for example, found only a weak relationship between farmer assessment of fraud levels and the actual quality of the herbicide glyphosate found in the local market. In fact, distinguishing counterfeits is even more difficult and, in most cases, requires dogs or lab testing, unless the fake is so poor. This conclusion was confirmed by the survey work conducted in Mali on this issue by Haggblade et al. [16], who took suspected counterfeits to the local registered importers for assessment. The results showed that importers could not tell, in some cases, whether the products were counterfeit or authentic from the label alone. To really know, they had to pull a sample from the bottle and test the chemical signature—not necessarily of the active ingredient, but of all of the co-formulants as well. Their labs could distinguish counterfeits readily from the spectrometry. However, simple visual inspection of the label is not always reliable, even for authorized importers. Besides lab testing methods such as biological and physiological indicators, there are chemical signatures, breakdown spectroscopy, and near-infrared spectroscopy [51,52]. Over the past few decades, a variety of digital anti-counterfeiting measures have been developed, from barcodes to holograms, radio frequency identification tags, and invisible pigments, inks, and infrared markers, and more recently, embedded nanotechnology-based solutions to ensure the full traceability and accurate detection of fraudulent products [9,53,54,55]. However, such techniques could be stand-alone systems that do not encourage collaboration among stakeholders in overcoming challenges such as developing a uniform protocol for inspection, surveillance, and monitoring, and using a platform to register pesticides [54,56,57].

To protect agribusiness against fraudulent pesticides, farmers are strongly advised by authorities to follow the recommended precautions to minimize the risks of buying and using these pesticides. Examining recognition level enabled an assessment of the extent to which respondents were practicing precaution recommendations. In particular, a medium variation in the respondents’ recognition of fraudulent pesticides was observed in the present study. Specifically, farmers’ recognition level had fallen to a low or moderate level for all of the practices investigated. This might be attributed to the farmers’ lack of knowledge about the importance of anti-counterfeiting measures as the main component of combating counterfeit and illegal pesticides on the market. Other possible explanations for this result could be due to the farmers’ inability to visually distinguish a legitimate product from a sophisticated fake copy due to the continuous improvement of counterfeit technology. The findings are in agreement with those of Zimba and Zimudzi [58], who argued that 22.7% of Zimbabwean farmers stated that they could distinguish between counterfeit and genuine pesticides to a large extent. Gharib [59] emphasized the need for awareness-raising and education activities for farmers on how to identify high-quality genuine agricultural inputs. Such programs may be particularly valuable for helping farmers make informed purchasing decisions when fraudulent pesticides exist on the market [26]. This education should also include a clear indication of who to contact for further information or a recommendation to report any suspicions in relation to fraudulent pesticides [60]. Globally, several industry-led initiatives have been piloted to address counterfeiting. The end-user authentication initiative conducted in Ghana is one example of verifying brand schemes [23]. In this initiative, CropLife (funded by Bayer) piloted the use of “Holospots” on Confidor, an insecticide for cocoa. Each container was marked with a hologram, which was verified by viewing under direct light and tilting the label. The user then texts in the numerical code shown to assess the authenticity of the product results.

The results of the binary logistic regression model indicated that the farming experience variable holding all other things constant is an important factor that influences farmers’ purchasing behavior regarding fraudulent pesticides. The variable of farming experience positively influenced their purchasing behavior and was statistically significant at the 1% significance level (p = 0.001). The coefficient of the farming experience of the farmers was positively associated with their purchasing behavior regarding fraudulent pesticides, which indicates the orientation toward non-purchasing among the more experienced farmers. An additional year in the farming experience of the farmers increased the log odds of not purchasing fraudulent pesticides (versus purchasing) by approximately 0.891 times (odds ratio = 891). Again, the farmers who had purchased fraudulent pesticides had a lower mean level of farming experience when compared to those who had not. A probable explanation for this finding is that more experienced farmers are more concerned about the health effects of fraudulent pesticides, as well as more aware of authorized and trusted pesticide shops compared to less experienced farmers. Previous studies have examined farmers’ pesticide use and safety behavior, providing similar evidence to our findings [61,62,63,64].

The variable of education significantly and positively influenced the purchasing behavior regarding fraudulent pesticides. A higher level of education was associated with farmers being more likely not to purchase fraudulent pesticides, in comparison to the group of less-educated farmers. When an additional year of education was attained by the farmers, the results showed that such farmers were approximately 1.2 time (odds ratio = 1.121) more likely not to purchase fraudulent pesticides (statistically significant at the 1% significance level (p = 0.00). Education is a helpful tool for farmers in analyzing the risks associated with the use of fraudulent pesticides, following the features and methods of recognition (visual inspection and analytical methods, as possible), and making decisions about the best options to purchase [17,60]. These results are in line with other studies that examined farmers’ behavior in pesticide use in general, such as the works by [29,31,65,66].

The results showed that the variable of farmers’ recognition of fraudulent pesticides was positively related to the likelihood of a farmer not purchasing such pesticides. The results in Table 2 show that those farmers who had higher recognition levels were more likely not to purchase fraudulent pesticides by approximately 1.795 times (odds ratio = 1.795) (statistically significant at the 1% level (p = 0.00) compared to farmers who had lower recognition levels. This shows that in terms of the non-purchase of fraudulent pesticides, as farmers’ recognition level changed in a positive way, they were more likely not to purchase such pesticides. These results confirm the findings of [12,16], who found that farmers’ adoption of recommended precautions during the purchase of pesticides had an effect on their fight against counterfeit and illegal pesticides on the market, suggesting that farmers’ recognition could be linked to their avoidance of the purchase of fraudulent pesticides.

Variable	Coefficients	Odds Ratio	Std. Err.	z	p > z
Constant	–25.734	0.000	0.000	36.384	0.000
Age	–0.315	0.730	0.572	0.320	0.572
Farm size	–0.004	0.996	0.088	2.916	0.088
Farming experience	0.116 **	0.891	0.001	11.741	0.001
Recognition	0.468 **	1.597	0.000	56.556	0.000
Off-farm income	–1.222	0.295	0.060	3.541	0.060
Education level	0.191 **	1.211	0.000	23.218	0.000
Extension as a source of information	0.083	1.087	0.887	0.020	0.887
Pesticide training experience	0.782	2.186	0.492	0.473	0.492
Perception of risks	–0.040	0.961	0.032	1.511	0.219
Chi-square (9) = 423.84 **; Nagelkerke’s R2 = 0.873; Log likelihood = 113.26

This study has certain limitations that should be taken into account. First is the inclusion of data only from a random sample of farmers from one governorate, which means that the study cannot be generalized to a national level or to other countries. Second, we analyzed their purchasing behavior using a dichotomous question (yes/no). This approach did not allow us to accurately measure the continuity of purchasing. In other words, we were unable to differentiate between a farmer who purchased fraudulent pesticides many times and a farmer who had purchased them only once. Therefore, future studies should include the intensity of purchasing during a specific period. Third, this study relied mainly on self-reports of farmers’ recognition levels. This led to an inability to assess whether farmers were able to correctly detect fraudulent pesticides. Thus, future studies should examine the relationship between farmers’ recognition level and the measured quality. Finally, many more factors such as subjective norms, beliefs and values, costs and benefits, behavioral control, and government policy are also worth investigating.

To the best of our knowledge, this study is one of the first to examine farmers’ recognition and purchasing behaviors regarding fraudulent pesticides. Our findings highlight that despite Egyptian farmers having a high perception of the health and environmental risks associated with fraudulent pesticides, the percentage of farmers who have experienced the purchase of such pesticides is high. Some of the key reasons for the use of non-genuine pesticides are difficulty in differentiating between genuine and non-genuine pesticides, the influencing power of distributors/retailers, and the low price of non-genuine pesticides. However, an in-depth examination of the influence of independent variables on farmers’ purchasing behavior indicated that farmers with more experience in recognizing fraudulent pesticides were more likely not to purchase these pesticides. In the same sense, non-purchasing behavior was positively associated with age and farming experience. In a practical sense, this study provides insights into the role that should be followed by farmers to tackle fraudulent pesticides, particularly during the purchasing stage. The policy implications of this study are multifold. Given that most farmers rely on pesticide retailers for information on pesticides, the government, along with different stakeholders, should organize training programs for the recognition of fraudulent pesticides. Moreover, developing an up-to-date list of all legally registered pesticides as well as implementing random inspection over pesticide distribution enterprises should be carried out. For farmers, awareness campaigns using new information technologies are required. These programs should aim to enhance knowledge among farmers about the negative consequences of using fraudulent pesticides, the main types of fraudulent pesticides, the features and methods of recognition, the authentication of the genuineness of pesticides through mobile phones, and how farmers can notify authorities of suspicious pesticides or activity. Due to the scarcity of research on farmers’ behavior with regard to fraudulent pesticides, more research in this area is needed to bring more diverse perspectives from different countries. This study shows the need for more comprehensive empirical research to estimate the amount of fraudulent pesticides in circulation rather than relying on a simple yes/no variable, by conducting a quantitative assessment to estimate the levels of fraud and disaggregating the results by active ingredient. Finally, suggesting an intervention framework for combating fraudulent pesticides in the context of agricultural knowledge and innovation systems could also be an interesting topic to consider.