Objective

I’ve been looking into minimum wage legislation and so I plan to use the institutional grammar approach and institutional analysis theory to analyze how minimum wage legislation (acting as an institution) effects behavior, choices and societal outcomes.

The main questions I want to investigate are:

  • why is there such a disconnect between what “boomers” (those who attended college circa 1970) understand as an acceptable minimum wage and a wage that is actually livable? It seems as if the idea of $15 is appalling to some people, but these people are engaged in economic sectors and aware of cost of living. So wouldn’t they understand why a 15 dollar wage is necessary? Or is that a narrative framed by policymakers in our realm only?
  • is minimum wage legislation designed for minimum wage workers in 2020?
  • what is the goal of minimum wage legislation? It’s been argued and framed as a living wage, but have the policies been designed, written and implemented to provide a living wage? Or is it just the narrative that its been framed in?

I tell a story examining minimum wage laws in the United States in order to assess and report on how populations who depend on minimum wage have fared over time. I want to examine how those effects change over time, (looking at the 1970s to 2020s), over jurisdictions (state and federal) and over space (states, or regions). Initially, I am examining Massachusetts and Texas, since Texas has always pegged their minimum wage to the Federal minimum wage, whereas Massachusetts has increased theirs. I want to know: Why are these wages so different across the country, time and jurisdiction for the same jobs?

I want to use the Institutional Grammar approach to analyze minimum wage legislation in order to investigate the intended policy outcomes and outputs as to compare that to the actual policy outcomes and outputs.

Audience Relevance

This story would be relevant to policy makers who are voting on/drafting minimum wage laws, academics studying minimum wage laws, and citizens who want to become more informed about the effects of increasing the minimum wage. Most importantly, the target audience is college students and young workers who are finding themselves making too little off of minimum wage to pay off loans.

Preliminary Data Analysis

Interviewing the Data

Below I will perform relevant data cleaning and aggregation techniques in order to get a sense of what the data are telling me. I am starting by just investigating the minimum wage rates in Texas, Massachusetts and the Federal government. This way, I can begin understanding what is worth comparing, and what questions will drive my search of literature. Note: literature here is referring to both a lit review, and the legislation that will inform my corups.

Secondary source: https://www.dol.gov/agencies/whd/state/minimum-wage/history# Primry source: FRED

Load the dataset

library(tidyverse)
## -- Attaching packages --------------------------------------- tidyverse 1.3.0 --
## v ggplot2 3.3.3     v purrr   0.3.4
## v tibble  3.1.0     v dplyr   1.0.5
## v tidyr   1.1.3     v stringr 1.4.0
## v readr   1.4.0     v forcats 0.5.1
## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x dplyr::filter() masks stats::filter()
## x dplyr::lag()    masks stats::lag()
library(readxl)
min_wage <- read_xlsx("min_wage_data.xlsx")

Preview the dataset

Here’s what the data look like:

head(min_wage)
Question #1

Is the frequency of the data conducive to further analysis or should it be aggregated/subet in any way?

yearly <- min_wage %>% 
  separate(Date, into = c("Year", "Month", "Day"), sep = "-") %>% 
  group_by(Entity, Year) %>% 
  summarise(yearly_average = mean(`Minimum Wage`)) %>% 
  filter(Year >= 1972)
## `summarise()` has grouped output by 'Entity'. You can override using the `.groups` argument.

This data is currently a monthly basis. This is not ideal, as minimum wage laws are unlikely to change on a monthly basis. I will start by finding the yearly average for each entitiy. Also, I only have data for all three entities after 1972, so I will subset the data to only look after 1972.

Question #2

How do MA, TX, and Federal Minimum Wages compare over time?

yearly %>% 
  ggplot(aes(Year, yearly_average, group = Entity))+
  geom_line(aes(color = Entity))+
  theme_minimal()+
  labs(y = "Minimum Wage (Dollars)")+
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

Texas has consistently been lower than or equal to the federal minium wage. The minimum wage in Massachusetts tracked very closely with the Federal minimum wage, but then began to diverge upwards in 1999

Question #3

In which years were the Massachusetts and Federal minimum wage equal?

MA_fed_same <- yearly %>% 
  filter(Entity != "Texas") %>% 
  arrange(Year) %>% 
  group_by(Year) %>% 
  mutate(diff = yearly_average - lag(yearly_average, default = yearly_average[1])) %>% 
  filter(Entity == "Massachusetts", diff == 0)
  
MA_fed_same <- MA_fed_same$Year

Massachusetts and the Federal Government had the same minimum wage in the following years: 1979, 1980, 1981, 1982, 1983, 1984, 1985, 1986, 1987, 1992, 1993, 1994, 1995.

Question #4

In which years were the Texas and Federal minimum wage equal?

TX_fed_same <- yearly %>% 
  filter(Entity != "Massachusetts") %>% 
  arrange(Year) %>% 
  group_by(Year) %>% 
  mutate(diff = yearly_average - lag(yearly_average, default = yearly_average[1])) %>% 
  filter(Entity == "Texas", diff == 0)
  
TX_fed_same <- TX_fed_same$Year

Texas and the Federal Government had the same minimum wage in the following years: 1988, 1989, 2002, 2003, 2004, 2005, 2006, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021.

Question #5

How many times did MA increase minimum wage? How many times did Texas? And the Federal minumum?

MA_increase <- yearly %>% 
  filter(Entity == "Massachusetts") %>% 
  arrange(Year) %>% 
  mutate(diff = yearly_average - lag(yearly_average, default = yearly_average[1])) %>% 
  filter(diff > 0)

TX_increase <- yearly %>% 
  filter(Entity == "Texas") %>% 
  arrange(Year) %>% 
  mutate(diff = yearly_average - lag(yearly_average, default = yearly_average[1])) %>% 
  filter(diff > 0)

Fed_increase <- yearly %>% 
  filter(Entity == "Federal") %>% 
  arrange(Year) %>% 
  mutate(diff = yearly_average - lag(yearly_average, default = yearly_average[1])) %>% 
  filter(diff > 0)

Massachusetts increased the minimum wage 19 times. Texas increased the minimum wage 5 times. The Federal government increased the minimum wage 17 times.

Reflection and Next Steps

How has your story idea changed after interviewing your dataset, if at all? If your story idea is changing after asking these questions, take note of that here. If it has not, what about your analysis to date gives you greater confidence in your idea?

After interviewing my dataset, I have generated a list of questions about the legislative side of this story that the data do not tell. Specifically, after looking at the wages of Texas over time in comparison, I wonder what defines the lag where the state minimum wage falls below the federal minimum wage. Are there rules defining when legislation can be updated? What obligation do the states have to update their minimum wage laws? Is this obligation defined by relative dates (relative to federal policy) or concrete dates (i.e. 5 years after federal updates, states must update). It also made us wonder how it is that Texas spent so long paying it’s workers less than the federal minumum wage. Why is Texas’ minimum wages so much lower than Massachusetts, and especially federal minimum wages? What is the process for creating and updating minumum wage legislation?

What else would you have to do in order to fully report out this story? For example, who might be some good human sources that could be integrated into this story? What other datasets or documents could aid you in reporting this story?

I need to collect information that will control for inflation, population, labor market inventory, cost of living, other wage legislation, other legislative bodies in order to draw conclusions about comparisons made between states. This might include historical CPI-U data, Census data, Bureau of Labor Statistics data, and minimum wage legislative documents for each geography.

In terms of human data sources, it could be very useful to add anecdotal evidence about those who are minimum wage workers, and from those who have a hand in designing and updating legislation. These human sources might look like:

  • A 2015-2020 college grad who earned money during college through a minimum wage job/worked a minumum wage job in general
  • A 1965-1979 college grad who earned money during college through a minimum wage job/worked or works a minimum wage job in general
  • A professor or legislator with expertise on the design and process of minimum wage legislation

Initial Exploration of Theory and Methods

Intro

As a student in the Data Analytics and Computational Social Science program, my research interests are driven by my love for empirical applications and based in my interest in institutional analysis. I was also a tutor at the Writing Center on campus during my undergrad years and have explored pedagogy which dissects linguistics and the social aspect of writing. Overall, I would like to use my data analytical skills to evaluate the impact of words in scenarios that use text-as-data, such as the Institutional Grammar Research Initiative. Given that I am currently enrolled in Prof. Brenda Bushouse’s public policy seminar, I would love to flesh out the theoretical assumptions of the IGRI’s mission in her class, and pair it with some of the analytical work in Prof. Bushouse’s class simultaneously.

This week I visited the IGRI website and navigated through all of the tabs. I took notes while I got myself acclimated to their mission, which is detailed in the sections below. Afterwards, I listed a few ideas of institutions that I could possibly analyze. I believe my choice is going to be student loans, with a goal of investigating the behaviors of students that might be influenced by promotions, gimics, and headlines. I think it would be interesting to compare the headlines/tag lines/quick reads to the fine print, promisory notes and repayment plans.

IGRI Mission and Research Program: Notes

https://institutionalgrammar.org

Core Foci: theoretical and methodological advancement of Institutional Grammar
Theoretical: PoliSci 780
Methodological: PoliSci 797TA

How? Computational Text Analysis: development and application of computational text analysis and supervised machine learning approaches for evaluating institutions, based on institutional grammar.

Evaluating Institutional Performance: utilizing institutional grammar to develop theoretically informed criteria for assessing the quality of institutions.

Using Institutional Grammar to study Simulated Behavior: interaction between formal and informal institutions, explore how institutional grammar can be used with game-theoretic approaches and agent-based modeling to facilitate institutional modeling analysis in silico

Institutional Analysis:
Institutions are constraints created by humans for the purpose of shaping behavior and interaction, i.e. public policy or social norms. Institutional analysis is the analysis of how the various kinds of institutions are created or designed, and how they affect human choices and societal outcomes.

Evaluating design: assessing content to understand structure, assessing information that content conveys about who can do what, when and how.
Evaluating impacts: assessing the implications of institutional design on behavior in practice, assessing implications of such on the attainment of institutional goals

Institutional Grammar:
The Institutional Grammar is a theoretically informed approach for organizing the content of governing institutions along generalizable features, in relation to which behavioral outcomes can be assessed. Institutions are comprised of individual behavioral directives and these directives typically convey common types of information: an action; an actor associated with this action; whether this action is required, allowed, or forbidden; the temporal, spatial, and procedural parameters of the action, and; rewards or sanctions for performing or failing to perform the action as prescribed.

  • proposed a linguistic syntax designates a unique syntactic element to each of these types of information.

Evaluating behavior in practice: organize content of institutional directives in accordance with syntax, systematically collect and analyze specific ways that institutions are intended to compel behavior, aggregated and syntactically parse institutional data to provide a comprehensive depiction of the scope of institutions.

Empirical Applications
Computational Text Analysis: Machine Coding of Policy Texts with the Institutional Grammar
Project Description: Applying the Grammar of Institutions in a particular research setting is resource- and time-intensive, precipitating concerns over whether it might ever enjoy widespread use. To address this, our team of scholars is developing automated approaches and open-source tools for coding institutions based on the Grammar. Preliminary work, applying insights from computational linguistics and natural language processing, was recently published in Public Administration as part of a special issue on methodological advances related to the Grammar. Currently, the team is continuing the development of the approach, while also preparing an R package to increase the accessibility of the Grammar for other scholars.

Institutional Performance: Procedural Tools for Effective Governance (PROTEGO): Patterns, Outcomes, and Policy Design.
Project Description: PROTEGO draws on a fundamental claim: the design and combinations of procedural regulatory instruments have causal effects on the performance of political systems, specifically on trust in government, control of corruption, sustainability and the quality of the business environment. The key mechanism in this causal relation is accountability to different types of stakeholders. PROTEGO provides a theoretical rationale to capture these accountability effects by adopting an extension of delegation theory that considers multiple stakeholders. Empirically, this research program collects, validates and analyses original data across the EU and its 28 Member States for the period 2000-2017. The new dataset maps in detail the design of the following administrative procedures: Notice and Comment, Freedom of Information, Impact Assessment of proposed legislation and regulation, Ombudsman procedures and Administrative Judicial Review.

Institutions and Behavior:Automating Policy Analysis using Agent-Based Modeling
Project Description: This project aims to develop an interface for the automated generation of agent-based models from policy specifications for complex domain models. The objectives of this project are threefold: (1) The implementation of complex domain-specific models to explore the usability of agent-based modeling as validation mechanism for empirical observations; (2) The refinement of the Institutional Grammar to afford computational tractability and facilitate the automated population of agent-based models with policy information; and (3) Analytical comparison of emerging formal and informal institutional configurations. Outcomes of this project include a) the refinement the encoding guidelines for the Institutional Grammar and the b) provision of a generic framework for policy analysis using agent-based modelling based on encoded policy information.

Initial Ideas:
- Social Distancing Guidelines
What guidelines produced what results?
Interpretations of ambiguous guidelines
Compared to NH?
- MA DUA
Confusing & disenfranchising guidelines
- Higher Education
Navigating as a first-gen student
Payment
- Loans
Read the fine print
Why do so many people purchase a bad loan (i.e. mortgage)
- Student loans
No defaulting
Sign at 18
First-gen student
Promos vs promissory notes