Rising Rice Prices?
A Longitudinal Study into the Interplay of Agricultural Goods
Despite the promises and reforms of the government aimed to lower costs of goods, the Filipino people still have not experienced relief from the continuous price hikes of goods. Thus, there have been movements in the government to review this law. However, a review of related literature suggests that existing studies that analyze price movement of agricultural commodities are lacking in either coverage or recency.
Overview
Problem
Despite being an agricultural country, there is a lack of research regarding price behavior of agricultural products. In recent years, prices of goods have been continuously rising, warranting an investigation behind the possible factors that affect it.
Solution
Hence, we used data science to gain insight into the trends and patterns of agricultural prices and subsequently learn the significant factors that affect prices of agricultural goods.
Research Question 1
Is there any notable correlation between the prices of different agricultural goods?
Research Question 2
Are there any seasonal patterns or events that affected the prices of agricultural goods?
Null Hypothesis
Prices are NOT affected by seasonal trends nor other agricultural goods.
Alternative Hypothesis
Prices of goods are AFFECTED by time of the year, as well as prices of other goods.
Data Collection
Source
We collected the data regarding retail prices for various agricutural goods from the public database of the Philippine Statistics Authority (PSA). They were downloaded as an excel (.xlsx) file.
Timespan
Dataset for monthly retail prices are from January 2020 to December 2023.
Count
In total, we have 201 various products with their retail prices spanning over 52 months, giving the dataset a total count of 10,452 items.
Data Exploration
Preprocessing
We computed the average of the Philippines as a whole, removing the null values that some regions had on certain products.
Visualization
We used Matplotlib and Seaborn to generate the graphs that showcased relationships within the data.
Correlation Test
We used the Pearson Correlation test to find correlations between two different agricultural products. We then used a correlation matrix to visualize correlations of multiple products.
Seasonality Testing
For each crop, we aggregated mean prices and calculated the z-score of their respective prices. We then performed cubic spline interpolation for better visualization. We made four line graphs to visualize seasonality.
Want to access our spreadsheet?
Click the link below to know more about the data we've dealt with.
Check out our spreadsheet!Curious to check our Jupyter Notebook?
Click the link below to check our exploration process.
Our Jupyter Notebook!Results
Top Products with Positive Correlation
Coconut and Corn
Beef and Coconut
Chicken and Beef
Top Products with Weakest Correlation
Rice and Ginger
Rice and Saba Banana
Chayote and Saba Banana
Top Products with Negative Correlation
Suaje Shrimp and Corn
Suaje Shrimp and Coconut
Beef and Ginger
Seasonality of Products
Animal products
Spikes in ber months and dips in the following months.
Seafood
Does not experience drastic changes in prices, but shrimp gets quite cheap during the ber months.
Fruits, Vegetables, and Roots
Does not follow some sort of general seasonality.
Rice, Grains, and Starches
Gradually increase in price from August up to December.
Discussion
Correlation of Prices
Establishing correlations between two products allows us to
Forecast Prices
When two products are correlated, a price change in one means a price change in the other, letting us forecast prices as one product experiences price changes.
Optimize Production
Diversifying the production of agricultural goods can protect our farmers from price movements that affect their possible profits.
Seasonality of Products
Ber months
The price increase during ber months observed throughout multiple products (seafood, meat, grains, rice) indicate higher demand during the holiday seasons.
Rice
Dry season rice is harvested during March and April. Hence, during the holiday season, its dwindling stockpiles, compounded with higher demands explains the gradual increase during ber months.
What's the significance?
By knowing correlations of products and their seasonality, we can
Help Farmers
Farmers can make informed decisions on what to plant in order to offset price movements. This reduces their risk of losing money.
Help Consumers
We can carefully predict and plan our purchases and potentially save money, especially in this time where prices of products and services are extremely high.
Conclusion
In conclusion, we reject the null hypothesis. Some of the
main takeaways
of our study include the following:
Price Movements
Prices of some products are affected by inter-product correlation and time of year.
Holiday Seasons
We have verified that ber months or the holiday seasons are time periods wherein prices peak for numerous products.
Applications
This study about agricultural products can help us, the consumers, and, most importantly, our farmers.
Some Recommendations
Due to the sheer number of agricultural products and the amount of possible product pairings, we were unable to investigate all of them. Hence, we recommend taking the time to perform correlation studies between all possible product pairs as well as using data that spans for a longer period.
