Andrew E. Jones Math 214-002 Lab 2: Barley Harvests Purpose: The purpose of this lab report is to compare several aspects of barley crops from 1972 to 1995 on a state by state basis as well as the entire US as a whole. By examining the differences from state to state and over the course of time we hope to notice any trends that may occur that would affect the production of barley. Introduction: The Department of Agriculture collects various statistics concerning crops and crop harvests each year. We will be looking at some statistics that were collected from 1972 to 1995 regarding the barley harvests for each year. Among the statistics we will be examining are ratios between amounts planted and harvested (in thousands of acres) and the number of bushels produced (in thousands). First we will be looking at the numbers concerning North Dakota’s (ND) barley harvest and production as it is by far the country’s most proficient producer of the crop. In 1987 the US saw a record amount of barley crops being produced, with ND alone producing over 25% of the total harvest. Though the amounts planted, harvested and produced vary and don’t always agree in regards to which state is the most proficient at growing, ND is consistently at the top of each category each year. To get started we will look at the raw statistics concerning ND’s harvests from 1972 to 1995. Descriptive Statistics: ND Harvest (in Thousands of Acres) Variable ND Harvest N 24 N* 0 Mean 2446.5 SE Mean 95.6 StDev 468.2 Minimum 1500.0 Q1 2110.0 Median 2450.0 Q3 2740.0 Maximum 3450.0 North Dakota’s mean harvest is a staggering 2,446,500 acres on average each year, which is almost 28% of the average total US harvest. This number is quite close to the median harvest of 2,450,000. This indicates a fairly consistent harvest coming from North Dakota, an assumption that is confirmed by the relatively small standard deviation of 468,200 acres (relatively small when concerning agriculture, that is). The median and standard deviation leads one to predict the grouping of harvests by amount should be fairly centralized, so let’s look at a histogram depicting the actual amounts. Histogram of ND Harvest 7 6 Frequency 5 4 3 2 1 0 1500 2000 2500 3000 ND Harvest (in Thousands of Acres) 3500 Andrew E. Jones The histogram of North Dakota’s harvest in thousands of acres matches our expectations. Namely, the number for the harvests are centered around 2,5000,000 acres, but that the graph has more of a peak than a sloping curve. The bars at either end of the histogram that portray extremes of harvests are quite small. This is exactly what you want to see if you are a farmer; it means that harsh periods of time for crop production don’t last extremely long and that barley prices will remain fairly stable. The slight increase around 2,000,000 acres is indicative of slower/harsher years resulting in this lower harvest, an idea that will be confirmed when looking at a plot of ND’s harvests over time. Time Series Plot of ND Harvest ND Harvest (in Thousands of Acres) 3500 3000 2500 2000 1500 1972 1976 1980 1984 Year 1988 1992 Andrew E. Jones When looking at the amount of barley harvested in North Dakota over the years you immediately noticed a drastic decline from 1978 to 1980. These corresponds exactly with a heat wave and drought that hit in 1980, one that was so severe it was published by the National Oceanic and Atmospheric Administration as one of the US’s billion-dollar weather disasters. The few years that followed this incident the state saw record highs for harvests, a number that I find perplexing as I wasn’t able to find information as to why. I suspect the government stepped in to assist farmers by improving infrastructure followed by a period that would have resulted in good yields without intervention, but this is mere speculation. After these record highs and lows the amount of barley harvested by North Dakota appears to stabilize, so let’s look at their actual production statistics. Descriptive Statistics: ND Production (in Thousands of Bushels) Variable ND Production N 24 N* 0 Mean 111107 SE Mean 7691 StDev 37678 Minimum 45150 Q1 85503 Median 104015 Q3 137003 Maximum 184250 Their average production of barley was approximately 111,107,000 bushels per year. Their median production of 104,015,000 bushels appears to be significantly different than the mean, an observation that is confirmed by a standard deviation of 37,678,000 which is proportionally larger than those seen in their harvests. On face value these numbers indicate that the histogram’s bars representing ND’s production will be less centralized than that seen for harvests. Let’s see if this hypothesis stands true. Histogram of ND Production 7 6 Frequency 5 4 3 2 1 0 40000 60000 80000 100000 120000 140000 160000 ND Production (in Thousands of Bushels) 180000 Andrew E. Jones As was guessed the bars for the production of barley in North Dakota are not as centralized as those seen for the harvest. The actual production of barley in bushels is still centered about the median and mean (specifically around 100,000,000 bushels), though the extremes seem to be in disagreement with the statistics concerning harvests. While this is quite interesting to see it is also quite easy to explain. In seasons of drought a farmer might still be able to harvest the land of his planted barley, but the actual production of it would drop as some of what he harvested would be unable to be sold as it would be dead. The same conclusion can be drawn of the other extreme. In years where a farmer harvests the same about of land he may get a better yield, and thus production, if the growing season is quite good. Let’s take a look at the amount of barley produced by North Dakota over the years. Time Series Plot of ND Production ND Production (in Thousands of Bushels) 200000 175000 150000 125000 100000 75000 50000 1972 1976 1980 1984 Year 1988 1992 Andrew E. Jones The production drop in 1980 coincides with the low numbers seen for harvest. However, the slight declines that were seen in the harvest’s time plot around 1974 and 1988 are much more pronounced for production. These drops, not surprisingly, also coincide with other billion-dollar droughts caused by heat waves. Looking at the plot as a whole reveals that actual production of barley is much more sporadic than that of harvests. This fact reveals a disparity between the amount harvested and the amount produced. Specifically, even though more land is harvested more barley is not necessarily produced as a lot of it may be unusable. Now that we’ve looked at the trends concerning the harvests and production of barley for a single state we should look at the statistics for the country as a whole. Descriptive Statistics: US Harvest (in Thousands of Acres) Variable US Harvest N 24 N* 0 Mean 8753 SE Mean 317 StDev 1555 Minimum 6279 Q1 7528 Median 8528 Q3 9726 Maximum 11974 The mean harvest for the entire Unites States is 8,753,000 acres. This number is astonishing considering North Dakota alone accounts for 2,446,500 acres. The median amount of barley harvested is 8,528,000 acres, which is close to that of the average. What jumps out at me, though, is a standard deviation of 1,555,000 acres. Such a large deviation implies that the amount of barley harvested in each state varies wildly. This seems to mesh with the sheer amount of barley produced in North Dakota in relation to the country as a whole. For confirmation we can look to a histogram representation of these numbers. Histogram of US Harvest 4 Frequency 3 2 1 0 6000 7200 8400 9600 10800 US Harvest (in Thousands of Acres) 12000 Andrew E. Jones The histogram peaks at two bars, representing a harvest of 7,8000,000 and 8,400,000 acres. The distribution shows that the amount of barley harvested each year is almost normalized. Presumably this is due to just a few Midwestern states being harvesting the majority of all barley. When those states have a bad harvest due to harsh weather it affects the entire United States, causing the bars of the histogram to be so close in height. Another observation is the distinct lack of a bar centered at 10,8000,000 acres. At first I dismissed this observation thinking that it only indicates that within the 24 years of statistics the combined total harvest within the Unites States did not fall within 10,200,000 and 11,4000,000 acres. However, this fact in of itself is fascinating. It represents a period of time where the number of acres harvest for barley skyrocketed by over 1,000,000, then suddenly plummeted down to a more normalized level. My prediction is that this coincides with the level of high harvest and production seen in 1984 to 1986, but I delay my justification until this is confirmed by looking at the time plot. Time Series Plot of US Harvest US Harvest (in Thousands of Acres) 12000 11000 10000 9000 8000 7000 6000 1972 1976 1980 1984 Year 1988 1992 Andrew E. Jones The amount of barley harvested by the entire United States does indeed peak between the years that I had predicted. My reason for predicting this surge of harvested barley is due to the drought that occurred in 1980. The heat wave that prevented a large harvest of barley during those couple years would have caused the prices for it to go up dramatically in the couple of years that followed. This fact would incentivize farmers to drastically increase the amount of barley they produce in order to sell what barley they could at such high prices. This would also account for the reduction in harvest in the years that follow as the prices would have dropped dramatically with a large influx of barley and subsequent price drop. Next we will look at the statistics concerning the US’s actual production of barley. Descriptive Statistics: US Production (in Thousands of Bushels) Variable US Production N 24 N* Mean 0 437956 SE Mean 17499 StDev 85729 Minimum 289994 Q1 380123 Median 421958 Q3 499580 Maximum 608532 The average production of barley for the entire United States is 437,956,000 bushels, 25% of which is from North Dakota which follows the trend we have already noticed. The median for production was 421,958,000 bushels, a number fairly close to the mean. The standard deviation of 85,729,000 bushels seems to be quite large, so I’m interested in what the histogram representing these numbers will look like. Histogram of US Production 9 8 7 Frequency 6 5 4 3 2 1 0 300000 350000 400000 450000 500000 550000 US Production (in Thousands of Bushels) 600000 Andrew E. Jones The frequency distribution for the production of the Unites States as a whole is particularly interesting. Ignoring the bar centered around 600,000,000, the observation would be that the histogram and by extension production of barley appear to be fairly normal. However, since we cannot simply disregard the outliers in this case those occurrences cause the entire histogram to be shifted. Given the previous trends for production and harvests it is safe to assume that these happened between 1984 and 1986, which will be confirmed when we glance at the time plot. Time Series Plot of US Production US Production (in Thousands of Bushels) 650000 600000 550000 500000 450000 400000 350000 300000 1972 1976 1980 1984 Year 1988 1992 Andrew E. Jones As expected the three years of record breaking barley production are 1984, 1985 and 1986. The years of low production that were previously seen in North Dakota’s time plots are also in the graph for the entirety of the United States which confirms our knowledge of there being a drought during those periods (although my cursory research on the subject had already acknowledged that). What is interesting to note is the apparent lack of stability in barley production over the course of time. The schema of production seems to always be headed towards a high or low point and rarely at a stable level. This pattern is also present when looking at just North Dakota’s production of barley. Since we’ve looked at both the production and harvest of barley, let’s now look at the relation between barley that is planted and that which is harvested. Scatterplot of Harvested vs Planted for California Harvested (in Thousands of Acres) 1100 1000 900 800 700 600 500 400 300 200 200 400 600 800 1000 Planted (in Thousands of Acres) 1200 Andrew E. Jones Picking a state at random, in this case California, reveals a linear relationship between the amount for barley that is planted and harvested. While this state was chosen for no particular reason, the linear nature holds true for every state. When more barley is planted, more is harvest. This makes sense as a farmer would want to harvest what they’ve planted in order to make as much profit as possible. One would expect, then, that if more barley was planted that more would be produced, but we know this is not the case. That expectation isn’t realized as droughts or other natural factors can kill crops. So even though those crops would be harvested a larger percentage of them would be unusable, hence hurting actual production. Conclusion: There are several things to note about this report that apply not only to barley, but of all crops. Heat waves cause drought and, no matter how much preparation is made, can obliterate crop yields. These periods tend to cause an increase in price for the crop, which is then over produced in the following seasons in an effort to maximize profits due to the price hikes. These droughts have long lasting effects on not just production of the crop but also the economics that surround them. These trends are most recognizable when looking at a time plot; they give the viewer a sense of how a couple bad years can lead to over production as soon as the drought is over. These trends aren’t recognizable when one looks at these statistics via a histogram. They only allow the observer to see where the median of production lie without a sense of time. Both of these graphs, along with the statistics that back them, are important tools for agriculturalists and economists alike.