Table 1.
Summary statistics for proposed and finalized rules, and the public commentary.
The reported statistics for public comments are averages over the number of proposed rules.
Table 2.
Ten most frequent words for different dictionary-based text measures.
Fig 1.
Number of comments received by the CFTC for proposed rules over time.
One rule that received over 9,000 comments is plotted at 2,000 for visual interpretability. The solid line is the local average computed using a loess smoother.
Fig 2.
Sentiment of comments submitted by interest groups for proposed rule 75 FR 3281, which introduced a number of new requirements for registration, disclosure, recordkeeping, financial reporting, minimum capital, and other operational standards with respect to retail foreign exchange (forex) transactions.
The rule was proposed on January 20, 2010, received over 9,000 comments, and was finalized on September 10, 2010 in the Federal Registar under 75 FR 55410.
Fig 3.
Timeline of rule-making at the CFTC for swaps.
Each event marks the introduction of a proposed rule in the Federal Register to the publication date of its final version.
Table 3.
Frequency of comments from different segments.
Fig 4.
Kaplan-Meier survival curves for two events: (i) rule proposal and (ii) rule finalization. For (i) event times in days are defined using the Dodd-Frank Act passage as day zero. For (ii) an event time is measured in days starting with the rule proposal date. Note that some rules are never finalized.
Fig 5.
Properties of proposed rules over time.
The left panel shows word count, central panel shows the percentage of highly litigious rules (defined as having a greater percentage of litigious words compared to the median level), and right panel shows the Wasserstein distance between the proposed and final rule. The solid line is the local average computed using a loess smoother.
Table 4.
Regression results testing whether proposed rules were shorter and more litigious during the first 400 days following passage of the Dodd-Frank Act.
A litigious rule is one when it has a greater percentage of litigious words compared to the median level and zero otherwise.
Table 5.
Cox proportional hazards model where the event time is number of days from the Dodd-Frank Act becoming law to rule proposal.
Table 6.
Negative binomial regression model estimation results, where the dependent variable is the number of comments submitted for a given rule.
Table 7.
Cox proportional hazards model where the event time is number of days from rule proposal to rule finalization.
“SM” is short for Social Media and “Fin” for Finance.
Table 8.
OLS regression explaining the amount change between the final and proposed rule.
“SM” is short for Social Media and “Fin” for Finance.