Volumes 6-7 (2 Jan – [31 Dec 1936]): “Gold Standard, Gold Reserve Act of 1934, Exchange Stabilization Fund, & Tripartite Agreement”¶

by RB, JB, JM Group SevenThreeTwentyTrio

Data Overview, Telling Our Story¶

Henry Morgenthau, Jr. served as Secretary of the Treasury for 12 years (1934 to 1945) under President Franklin D. Roosevelt. During his time as Secretary, the United States Department of the Treasury implemented many monetary policies to help stabilize the national and international economies affected by the Great Depression. The Gold Reserve Act of 1934, the Exchange Stabilization Fund, and the Tripartite Agreement of 1936 were notable policies that would have a lasting impact.

After the decision to remove the U.S. dollar from the gold standard and allow depreciation against foreign currency in 1933, the Roosevelt administration sought to stabilize the exchange rates, which would, hopefully, stabilize domestic prices (Henning, 1999). The desire for stabilization led to the passing of The Gold Reserve Act in 1934 (Henning, 1999). The Gold Reserve Act, also called The Gold Act, is a law that transferred the gold and gold certificate titles from private individuals and institutions to the United States Treasury (Hayes, 2022). This act stopped banks, financial institutions, and the Federal Reserve from exchanging dollars for gold (Hayes, 2022). It gave the government more control over the money supply by allowing the Treasury to buy gold internationally to devalue the U.S. dollar in foreign exchange markets (Hayes, 2022).

The Exchange Stabilization Fund (ESF) was created under Section 10 of the Gold Reserve Act in 1934 and renewed for one year by President Roosevelt's proclamation in 1936 (Henning, 1999). Per Section 10, the goal of the fund was to stabilize the exchange value of the dollar, giving the Secretary of the Treasury authority to deal in gold, foreign exchange, and other instruments of credit and security (Hennings, 1999).

One result of the ESF was the Tripartite Agreement of 1936, which occurred in September 1936 between the United States, Great Britain, and France. The Morgenthau Press Conference Papers, Volume 7 table of contents describes it as an “original agreement between [the] United States, England, and France” (FDR Library, n.d., p. 7). This agreement pledged the stabilization funds of the three countries to maintain gold value and currency exchange rates (Henning, 1999). Similar agreements between the United States and other countries occurred in 1936 and subsequent years.

Morgenthau spoke to the press on numerous occasions in his role as Secretary of the Treasury, answering frequent questions about the Gold Act, the Fund, and the various agreements between other countries. These hundreds of press conferences were transcribed and microfilmed and are housed in the Franklin D. Roosevelt (FDR) Presidential Library & Museum. [JM]

Dataset Exploration related to Scale and Levels within Collection¶

Our Volumes, 6 and 7, of the press conference transcripts started on January 2, 1936, and ended on December 31, 1936, with a total of 799 pages (including some accidental duplicates created by the originators of the documents).

We conducted three Zoom meetings and created a group chat to discuss the project. Based on the research that was conducted, we decided to focus our narrative on monetary considerations. The 1936 press conferences discussed monetary resources – specifically gold and silver – and international policies regarding these resources. [RB]

Data Cleaning and Preparation¶

We started the data cleaning by testing different Optimal Character Recognition (OCR) tools. We entered the PDF of Volume 6 into DocDrop to redo the OCR, and then copied and pasted the text from the PDF into OpenRefine with mild success regarding the readability of the text. We also signed up for a 7-day trial of Adobe Acrobat Pro and attempted to use the scan/OCR function to enhance the scanned files, but the data were still illegible. Ultimately, we split the two large PDFs into smaller, more manageable chunks, creating one PDF for each press conference. Each group member took one-third of the PDFs and ran them through OCR Engine 2 (Figure 1). After copying and pasting the text into Word documents, we used the “Fix any mistakes” option from Grammarly to clean the text further, and then manually checked the PDFs for accuracy. Each group member spent an average of 12 hours transcribing the press conference papers, with 35 PDFs per person and an average of nine pages per PDF. [RB]

During our data preparation process, we researched two programs that utilize AI: Google Notebook LM and Grammarly. We considered using Google Notebook LM to review data for the project, and though initial results seemed promising, we noticed inaccuracies while using the AI tool. The creators of Google Notebook LM claim that the AI only utilizes the sources provided by the user; however, the word “Administration” was assumed to indicate the “Biden Administration” in the document summary. Our two volumes do not mention President Joseph Biden, as he was not born until 1942 and our Morgenthau volumes covered 1936. Due to the inaccuracies in the AI tool, our group pivoted to OpenRefine for data review. [JM]

The second tool, Grammarly, is described as a “free AI writing assistant.” During the data cleaning process, we used a function called “Fix any mistakes,” which rewrote entire portions of text with surprisingly accurate results. In this case, “accurate” means “like the original” and not necessarily “grammatically correct” which was important because we wanted to review what was written by the transcriber rather than what would be considered “correct” today. [JM]

Modeling: Computation and Transformation¶

After using OCRSpace and Grammarly to ensure our dataset was legible, we copied the cleaned text and pasted it into the Clipboard function of OpenRefine. This process resulted in 6813 rows. The “Text Filter” function was selected to analyze the dataset and determine the most popular terms within the Volumes. The way our dataset was organized made it so that there were only unique rows of text, which is why we selected the “Text Filter” (Figure 2) We chose to use the words from the Table of Contents at the beginning of each volume as our beginning glossary. Additionally, each group member noted words of interest or high frequency when cleaning their portion of the dataset. We also used “CTRL F” (the “Find” feature) for a more accurate word count, in case there was more than one mention in a single row of the dataset. Also, to enhance the accuracy, we searched alternate terms that essentially indicated the same topic, for example, “Germany” and “German” or “Resign” and “Resignation.” After compiling a list of words and their frequency within the dataset, we grouped the words together under topics, such as “Countries” and “Money.” Creating these groups helped with the creation of the visualizations. [JM]

From our OpenRefine search, key topics emerged, including the gold standard; effects from The Gold Reserve Act of 1934; interest in the stabilization of domestic and foreign economies through the Exchange Stabilization Fund; and the Tripartite Agreement between the U.S., Great Britain, and France. Our primary focus remained on monetary considerations during 1936.

Modeling: Visualization¶

As mentioned previously, we first determined how many times words from each Table of Contents were in the text of the press conference transcripts. We chose to display this information in a word cloud which uses size and color to indicate the frequency of words (Figure 3). [JM] Note: The word cloud depicts all terms we searched for in OpenRefine for our research. The size and color of the word indicate how often it was mentioned. Large, gold words are mentioned most often, whereas small, silver words are mentioned much less.

People: We further divided the terms into groups to enhance our understanding of the press conference’s themes. Many people were listed in the table of contents of each volume, so we entered each name into the “Text Filter” of OpenRefine to determine how many times they were mentioned. Other than Morgenthau and members of his team who were recorded as helping him answer questions, and the mostly anonymous press members, the data showed important appointments and resignations were important press conference topics. For example, the appointments of Joseph Murphy and Frank Wilson as Assistant Chief of Secret Service and Chief of Secret Service, respectively, resulted in the first and third place of most mentioned people in Volumes 6 and 7 of the press conference transcripts. Resignations included that of Thomas Coolidge from his duty as Under Secretary and “Chip” Robert from his duty as Assistant Secretary. “Rumored” appointments to these positions were about Eugene Black and Wayne Taylor. The second most mentioned person was former President Herbert Hoover. We decided to depict notable people’s mentions via word cloud (Figure 4). [JM] Note: The word cloud depicts people mentioned throughout the press conference transcripts. Just like with Figure 1, the size and color of the words equate to how often they were mentioned. [JB]

Monetary Resources: Gold was one of the most frequent mentions within our volumes (see Figure 4). Because the FDR Administration first devalued the U.S. dollar in an effort to help domestic and international economies, then enacted the Gold Reserve Act in 1934, discussions regarding monetary resources and stability were frequent between Morgenthau and the press.

Countries: The United States Department of the Treasury interacts with many countries. We noted all the countries included in the table of contents of both volumes plus any other country we noticed when cleaning the data. Because interactions between the United States and other countries are a crucial topic, we made a bar graph to depict how often each country was mentioned (Figure 5). For increased accuracy, we counted multiple terms and tenses referring to the same location (e.g., “France” and “French,” or “Holland” and “Netherlands”). [JM] Note: The bar graph illustrates the number of times specific countries were mentioned in the press conference transcripts. [JB]

Tripartite Agreement: The Great Depression led to many decisions by President Roosevelt and the Department of the Treasury that affected the United States economy and other economies worldwide. The Gold Reserve Act, which created the Exchange Stabilization Fund, led other countries to consider and implement stabilization funds. To help stabilize exchange rates between currencies, the United States, Great Britain, and France entered the Tripartite Agreement, which was discussed earlier in our report. This agreement is depicted with a radar chart (Figure 6), showing the three countries in the agreement and how many times “tripartite” was mentioned in relation to those countries. France was mentioned the most often because they had not yet devalued their currency as Great Britain and the United States had. They did so to enter into the agreement. [JM] Note: The radar chart illustrates the three countries originally in the Tripartite Agreement of 1936 (United States, Britain, and France). The black line indicates how many times each country was mentioned with “tripartite” during the Text Filter manipulation in OpenRefine. [JB]

Creating the Visuals: We created four kinds of visuals: bar graphs, donut charts, radar charts, and word clouds. We selected the color scheme of gold, silver, and black to reflect the theme of money. After deciding on the colors, we saved the hex codes so the colors would stay consistent for easy transfers between Excel and the website https://www.wordclouds.com/ (a free site that allows you to create word clouds). We created the charts using Excel and used a 3D style as part of our theme. Bar graphs focused on the information that required reporting exact numbers. Donut charts focused on representing information that did not need the same accuracy but rather visual representations. Word clouds were used to convey the importance of words among a certain category. We used a radar chart to plot information on the Tripartite Act since it was an important part of our narrative but had very little information to chart. The donut chart and word clouds leave the visualization open-ended for the reader to further research. [JB]

Ethics and Values Considerations¶

We aimed to complete our work following the Society of American Archivists’ Code of Ethics and Core Values. We exercised professional judgment while working with primary source materials, promoted the use and understanding of the historical record, and recognized the importance of primary source materials to examine past events and gain insight into human experiences (SAA, 2020). [RB]

Summary and Suggestions¶

Our primary focus was on the monetary considerations of the gold standard, the Gold Reserve Act, the Exchange Stabilization Fund, and the Tripartite Act which Henry Morgenthau, Jr. discussed during press conferences in 1936. However, we noticed a significant number of phrases such as “off the record,” “just for background,” and “don’t quote me” throughout the transcripts by Morgenthau, as depicted in Figure 7 above. Note: The donut graph depicts how often terms such as “off the record” and “for background” occurred in Volumes 6 and 7 of the press conference transcripts, which felt like a notable pattern. [JB]

It appears Morgenthau and the press had a cordial relationship, and we wonder if the press respected the “off the record” requests or printed stories with details that were supposed to be off the record. An idea for future research is to review these mentions and compare them to newspapers printed at the time and place of the press conferences. We assume these would be Washington, D.C. papers, but further research is needed to determine the members of the press present at the press conferences, in addition to the presses they represented. Though there are some names mentioned throughout the transcripts, it is difficult to determine if these are names of members on Morgenthau’s team or members of the press. Further, future research could evaluate the ethical and moral implications of crafting a narrative by providing special information to the press but asking that it be kept from the public via phrases such as “off the record.” [JM]

In addition to our notable mentions, several words with high frequency of use, but that did not align with the focal point of our narrative, could be considered for further research, such as "bonds." Bonds were mentioned over 200 times, but with many kinds of bonds, it was impossible to include all of them in our narrative. Focusing on foreign policy, foreign bonds were more relevant to our project, however, they were only mentioned four times. Another suggestion for future research could be to investigate the types of bonds and their impact on the national and global economy in 1936. [JB]

With RB's ideas, JM's grading techniques, and my visualizations, we made an excellent group! [JB] I appreciated being part of a team where each member was eager to do their part and expressed the question of, “Am I doing enough?” to which the answer is a resounding “yes!” [JM]

References¶

• Franklin D. Roosevelt Presidential Library & Museum (FDR Library). (n.d.). "Press Conferences of Henry Morgenthau, Jr., 1933-1945." http://www.fdrlibrary.marist.edu/archives/collections/franklin/index.php?p=collections/findingaid&id=536
• Hayes, A. (2022, June 29). “Gold reserve act of 1934: Meaning, history.” Investopedia. https://www.investopedia.com/terms/g/gold-reserve-act-1934
• Henning, C.R. (1999). The exchange stabilization fund: Slush money or war chest? Institute for International Economics. https://www.piie.com/bookstore/exchange-stabilization-fund-slush-money-or-war-chest
• Society of American Archivists (SAA). (2020). “SAA Core Values Statement and Code of Ethics.” https://www2.archivists.org/statements/saa-core-values-statement-and-code-of-ethics