by NH, BM, MS
Our main goal was to analyze trends in topics and keywords chronologically. Thus, we first needed to look at the data in chronological order. While the bulk of the text was in chronological order, it could take far too long to analyze every page. Thus, an idea arose to put a primary focus on the index of each book. However, the indices were in alphabetical order. Fortunately, each entry included a date, allowing us to manually make rearrangements of the information.
Additionally, since tools like OpenRefine and Voyant struggled to read the data from the original PDFs, our next objective was to find a way for these tools to easily process the text. The whole book would take far too long to transcribe, and the index already contained a summary of when important terms and names were mentioned in the full text, so it seemed a good place to start. After that, analyzing and manipulating would be much easier. Thus, transcriptions of the indices were created.
Now, we had the topics in chronological order, but words on a page separated by dates were difficult to compare. Thus, the stacked bar charts were created to easily visually compare the trends of topics mentioned during the press conferences. Voyant can create a trend chart, but it has flaws. It separates the document into 10 parts, but not all parts may be equal - for example, there were six conferences in June, but only three in May. The bar charts work to balance the different amounts of conferences by converting the raw numbers into percentages per month. This way, a topic being mentioned twice in May (66% of the time) will be weighted differently than twice in June (33%).
Over all 3 volumes, certain terms are found to be extremely frequent throughout, such as financing, bonds, and war. The frequency of most terms fluctuates over time, like compulsory savings, which only came up between June and September. Such fluctuations likely occur as topics fluctuate in importance, with current events and policies determining the topics of discussion (BM).
During the year 1942, covering volumes 20-22, the war effort really started to take off (see Figure 1). Pearl Harbor had just occurred in December of 1941 and the United States had officially declared war and shifted our economy to what FDR called “the Arsenal of Democracy”. In 1942, federal spending increased by 132.15% compared to 1941’s 37.28% (Tassava 2008), in order to support this new war economy. In order to fund this new war economy, War Bonds were introduced in May of 1941 and really took off in 1942. The War bonds not only helped to fund the war effort but also made civilians feel like they were able to make tangible economic contributions to the effort. This was particularly important as more men enlisted or were drafted as the draft expanded to include men between the ages of 18 to 37 in November of 1942 (History.com 2024).
Taxes were another way that the United States helped pay for the war effort and personal property tax in particular. In 1942, the government introduced the new Revenue Act which lowered the personal income exemption from $1,500 to 1,200 (Janey 2019). With this new act in place, the government was concerned that Americans wouldn’t be too keen on paying more taxes, however, with the help of a beloved animated character, Americans understood just how important paying their taxes was in addition to buying war bonds. The Walt Disney Company was employed by the United States government to create a cartoon with Donald Duck happily paying his taxes in order to defeat the evil Axis powers. The cartoon ended up working and in conjunction with the continued purchasing of Defense Savings Bonds and Stamps, which was later renamed the War Savings Bonds and Stamps (p. 259 [Vol 21]), the American people supported the war effort by paying their taxes as well.
The impact of the Donald Duck cartoon did wonders for the financing of the war, with the Treasury Department receiving praise for their advertisements (p. 183-184 [Vol 20]) while only paying an out-of-pocket expense, unlike what the department would be earning through advertising freely over the radio or newspaper (p. 133-43 [Vol 20]). Despite the power of the animated duck, Morgenthau pushed for more taxation with the use of a study breaking down the changes in spending for different classes of American families, utilizing “purchasing power” of the average family unit (p. 260-281 [Vol 20]). This study additionally pushed for the “luxury tax” tagline added to qualified “semi-luxury” items such as cigarettes, candy, and soft drinks (p. 269-270 [Vol 20]). The voluntary purchasing of Defense Savings Bonds and Stamps promoted what the average American could do to help their country in the war, but there is always more that can be done.
Though the average American family wasn’t the only group facing additional taxation, Morgenthau unveiled his proposition for taxing outstanding tax-exempt state and municipal securities, and though the concept had been approved by the President with the account of the war, the Courts still held the right to contend with the taxation between city, State, and Federal Government (p. 57-63 [Vol 20]). Morgenthau’s plan had no intention of breaking any contract with “holders of outstanding tax-exempt government securities” with the creation of a taxable $20 billion of State and municipal securities (p. 73-80 [Vol 20]). This plan was met with only two telegrams, one in favor and one against (p. 94-95 [Vol 20]), the public had seemingly made up their mind on the additional taxation of the government.
With the war on everyone’s minds, so came the question of what to do with acquired assets from said war. Materials captured from Germany were set to be sold and added to frozen assets for the United States (p. 13-14 [Vol 20]), with the $7 billion of assets being kept and retained intact to “rid them of their Axis influence”, provide additional aid to the war efforts from the Allied side, and until they could be “disposed” of under negotiated peace terms (p. 168-71 [Vol 20]). This was a large increase in gains since the last world war, with 500 million assets being acquired during that time (p. 168 [Vol 20]).
The case for stabilization in the world was a large trend during the early months of 1942, starting with the introduction of a common unit of currency for international export trade, supported by a gold or silver stabilization fund (p. 41-50 [Vol 20]). This move was ultimately placed in a positive light, with the concerns over dealing with depreciated currencies; which later led to the notion of creating a North-South American Bank to stabilize the purchasing power of Americans in the neighboring Americas, where the currency exchange rate had been stopping business (p. 185-188 [Vol 20]). The plan was to expand the markets starting with the Americas and with the goal of branching outwards from there, to broader and wealthier shores.
Stabilization through the American hemispheres wasn’t the only use for the stabilization fund, as China was brought into the mix with worries it may join the Axis powers in the war. Morgenthau approached Congress with the proposition of a $500 million dollar loan to the country, with England providing additional aid, as a means to persuade China from becoming involved. The loan was called a “morale builder” for the developed country (p. 107-145 [Vol 20]) and was awaiting approval of Chiang Kai-shek as of March of that year (p. 198 [Vol 20]). This loan was subsequently taken out of the stabilization fund to accomplish the feat.
In addition to a common currency, the conversation shifted to the production of silver being mined by the United States, Morgenthau was determined to repeal the Silver Purchase Act and selling surplus silver to industries, where it could replace the quickly diminishing supplies of lead and tin (p. 101-106 [Vol 20]). His push to sell silver returns again over a month later, as he tried to find a method of releasing silver stocks without legislative authority, once again promoting the sales of excess silver as it had been already used to additionally replace copper and processing aluminum (p. 323-328 [Vol 20]). Morgenthau was sure that silver from the United States would easily meet, and potentially beat, global market prices, from Mexican and Canadian prices of 35 cents per ounce compared to the United States 71 cents an ounce (p. 398-403 [Vol 20]). Silver mined from the United States was an untapped market that would only raise the overall price of silver, leading to an increased revenue without wasting the silver the nation already utilizes.
1942 also saw a sharp decline in unemployment that continued to last until the end of the war in 1945 with just 2.7% of the population listed as being unemployed compared to the previous year of 5.6% (Tassava 2008). This also included the government forcing automobile manufacturers to convert to the production of planes for the war effort as well. Despite the overall support of the war from the American people, some aspects were not as positive. For example, taxes increased in order to provide for the war effort which made Americans nervous and can be seen in some of the questions that the press asked Morgenthau. While Americans were not completely forced to ration as some of the Allied European nations were, great sacrifices were made in order to provide for the war effort (FDR Radio Address 1942).
October’s press conferences included one run by Mr. Sullivan instead of Morgenthau. The majority of these talks were over war bonds, specifically how well they were doing at this time, as well as the introduction of new types of war bonds at a 2 percent interest rate. Reporters also asked a lot about salary limitation law, specifically a section restricting salary increases, and another which stated that no corporation can make a deductible item from taxes or salaries over and above twenty-five thousand.
November had seven press conferences, with two in a single day due to a delay in obtaining the correct numbers for the press to report on. One main theme was foreign affairs, specifically an expansion on French Frozen Funds with the new declaration that any French territory in Continental Europe should be considered enemy territory. Another theme was the fear that many Americans may not be able to pay their personal income taxes with new legislation, the Revenue Act, which lowered the personal income exemption from $1,500 to $1,200. It was also announced in the November 19th press conference that $9 billion would be borrowed in order to fund the war effort, one of the largest sums that the Treasury had ever borrowed before.
December only had four press conferences in 1942. Those press conferences were mainly about providing the last numbers for the year. Said numbers were primarily of war bonds and included reports about the first war bond drive that took place from November 30 to December 23rd, 1942. During this period they raised eleven billion, nine hundred and forty-six million dollars with sixty percent from non-banking sources. At the beginning of December, they also discussed an offer to lend silver to other countries on an ounce-for-ounce basis for coinage purposes.
In addition to the content of the press conferences, there were also additional themes that were noticed throughout the volume. The first was that Morgenthau would refuse to answer some questions asked of him by the press. Most of the questions were about other departments or individuals that he did not feel comfortable speaking about. He also had to dispel a lot of rumors or figures that the press would present in their questions, happening on a consistent basis. Speaking off the record was also quite a common occurrence when he did not wish to be quoted about something but still wished to deliver information to the press. Most of the language used in these press conferences was economic in nature since he was the Secretary of the Treasury. However, some terms, phrases, or jokes would be used that may not hold the same meaning today. For example, wildcat was used several times in the December 17th press conference, and while the meaning is not known for sure, the context of the phrasing indicates sleazy dealings (NH, MS).
Our main goal was to analyze trends in topics and keywords chronologically. Thus, we first needed to look at the data in chronological order. While the bulk of the text was in chronological order, it could take far too long to analyze every page. Thus, an idea arose to put a primary focus on the index of each book. However, the indices were in alphabetical order. Fortunately, each entry included a date, allowing us to manually make rearrangements of the information.
Additionally, since tools like OpenRefine and Voyant struggled to read the data from the original PDFs, our next objective was to find a way for these tools to easily process the text. The whole book would take far too long to transcribe, and the index already contained a summary of when important terms and names were mentioned in the full text, so it seemed a good place to start. After that, analyzing and manipulating would be much easier. Thus, transcriptions of the indices were created.
Now, we had the topics in chronological order, but words on a page separated by dates were difficult to compare. Thus, the stacked bar charts were created to easily visually compare the trends of topics mentioned during the press conferences. Voyant can create a trend chart, but it has flaws. It separates the document into 10 parts, but not all parts may be equal - for example, there were six conferences in June, but only three in May. The bar charts work to balance the different amounts of conferences by converting the raw numbers into percentages per month. This way, a topic being mentioned twice in May (66% of the time) will be weighted differently than twice in June (33%) (BM).
The first thing we did after splitting the volumes amongst the three of us was to input our respective volumes into OpenRefine. This was done multiple different ways but with very little results. We noticed that any program did not recognize the majority of the text in each volume so we attempted to then transform the PDF into a readable format. This also proved to be futile as each attempt was unsuccessful. For example, one group member attempted to use DocDrop with the original PDF but this conversion was unsuccessful and communicated to the rest of the group. Another group member also tried inputting the original PDF format into Google Notebook LM, but similar results were also found and communicated to the group.
After we felt we had exhausted our options with the original format of the PDFs, we first attempted to convert our respective Volumes into JPG files that could then be transformed into readable text, but again each program that we attempted to use did not recognize the text. This included an attempt where we attempted to use PrePost SEO which converts JPG files into machine-readable text and produces three pages at a time. This venture was not continued since the files required a lot of corrections and were decided to be too time-consuming. We also attempted this same process with the free OCR converter and Convertio programs which could only transform one page at a time, and once again, required many corrections. The closest we came was converting the now JPG file of a volume into a TXT file, however, most of the text was unreadable and lots of corrections would need to be done in order to have the proper text machine-readable.
It was at this point that we then decided it would be best if we could type out the index of each volume and see if this would be workable. Thankfully this did work and we were able to input the index from each volume into Open Refine, Voyant, and Google Notebook LM for testing. The index is not however a full story of the text so notes were taken on each press conference in the volumes and were input with the index, sorted by date, and run through Open Refine, Voyant, and GoogleNotebook LM. Once those were all successful for each volume, we gathered the top keywords that were used in each volume and the volumes were able to be analyzed (MS, BM).
Upon the initial download of the raw, photocopied material, attempts to place the digitized documents into data editors proved nearly impossible. Our team decided to transcribe the indexes into a text document, convert it into a data-based readable format, and were then able to upload the data into different data processing applications to discover recurring trends and create visual representations of our findings (NH).
The below line-and-bar graph (Figure 2) was created by Voyant from the index as-is, in its original alphabetical order.
We realized two important facts from Voyant’s trend graph. One: the graph is sectioned out automatically into 10 amorphous parts, and two: the results would have words lumped together, as can be seen in sections 1, 2, and 10. Thus, we rearranged our transcriptions into chronological order, using the dates attached to each entry. The following chart (Figure 3) more accurately represents the word trends over time.
The four charts below (Figures 4-7) were created from the transcribed indices, with Figure 7 compiling the three volumes’ results. The frequency of the topics of each month is compared, so one can see how the composition of topics changed over time. The chart shows the percentage of times that topic was brought up in conferences that month, each topic given a percentage from 0 to 100. Topics only mentioned once or twice are excluded.
Trends can be seen and analyzed. Topics such as government financing are constantly discussed, while silver is regularly discussed, but not frequently. You can also see how some topics are only discussed for a short period, and then fall away, such as national defense programs and stabilization (MS, BM).
Figures 8-10 present a more summative view of each volume’s most frequent words (bulleted and visualized), alongside our methods used.
When it comes to our work on the volumes, it is important to consider that since we were unable to translate word for word, this is not a complete view of the Press conferences in each volume. In order to truly get an accurate picture of the complete volumes, we suggest that a translation word for word would be the most accurate representation of these volumes, rather than an overview. This would be tedious, however, we think this would paint a more accurate picture, and allow for disability accommodations to be added such as text readers for visually impaired users.
We also noticed that several pages in these volumes were duplicates and they would need to be removed in order to again provide a more accurate view of the volumes. Or, if removing these duplicates would harm the integrity of the volume, it would need to be noted that duplicates are present in the record. Additionally, several words found in these volumes are phrases that are either racially insensitive or are no longer used today, and providing a warning would be a consideration to possibly add.
The chart provided takes its information solely from the index. If a topic was mentioned, but nothing noteworthy was said about it, it was not included in the index. Any topics discussed frequently but never spoken of at length may be underrepresented in the visualizations, which were made from index data (MS, BM).
Through the course of the data manipulations, Microsoft Word and Google Documents were utilized in transcribing the notations and generalities mentioned in the index to best comprehend the events composed in Volumes 20 through 22. Creating a readable document led to a smoother transition of recreating our findings in graphs and tables through OpenRefine and Voyant, which led to the understanding of key terms and themes throughout the volumes.
Looking to the future, the issue of the legibility of the PDFs should be expanded upon with concerns over the values of the document as a whole. An additional digital text document would allow for patrons unable to read the photocopied PDF to use assistive devices to engage with the materials virtually. That may be time-consuming to retype everything, but the benefits of allowing all users to utilize the information would open the documents to all as opposed to just those who can read it on their own.
Another addition to the documents would be additional contextual information commenting on the outcomes and effects of the different notions communicated in the press conferences. While that is additive to the document and will require further research and citation, it would give readers an understanding of both Morgenthau’s work behind the scenes and the public’s response, separate of Morgenthau’s interpretation of the American’s response.