1]: Develop a plot showing the kilograms of masks imported each month from China and Vietnam (aggregate across these two exporters) by water and air. Devise a second plot that shows the percentage of masks that were imported by airfreight. What patterns do you observe, and speculate why these patterns exist? (use Air SWT (kg) and Vessle SWT (kg) columns for this question)

[2]: Develop three charts showing the import districts through which masks arrived from China and Vietnam (aggregate across these two exporters) for the second quarter (April - June) of 2019, the second quarter of 2020, and the second quarter of 2021. Show imports based on both water and air freight and include in the charts the top 10 import districts by total weight of imports (Vessel SWT + Air SWT) for the quarter in question. How have the patterns evolved over time?

(use same columns as Q1, but by District)

[3]: There has been substantial interest in understanding the evolution of transportation costs for importing items such as masks. Focusing separately on the import districts of Los Angeles, CA and New York, NY, first develop a plot showing the insurance and freight cost per metric ton (kilograms divided by 1000) of total imports through these two locations (i.e., do not aggregate across the districts). Focus only on imports from China to rule out country-specific factors. What pattern do you observe in the insurance and freight costs per metric ton?

(use Freight Costs (column H) for this question)

[4]: A second measure often used by economists to represent freight costs is what is termed ad valorem insurance & freight costs, which is calculated as insurance and freight costs divided by the customs import value. Repeat the analysis done for #3 using ad valorem insurance and freight costs (note, use Column F's statistic for customs import value in the denominator). What pattern do you observe? How does this differ from what was observed in #3?

(use Customs Value (CIF Data) (column F) and Freight Costs (column H) for this question)

[5]: Devise a data-driven explanation that can reconcile the different conclusions reached in #3 and #4.