Important: This is a template repository to help you set up your team project.

You are free to modify it based on your needs. For example, if your data is downloaded using multiple scripts instead of a single one (as shown in \data\), structure the code accordingly. The same applies to all other starter files—adapt or remove them as needed.

Feel free to delete this text.

The Telecom Merger Effect: Market Share & Consumer Shifts

This project analyzes the effects of the 2020 merger between T-Mobile and Sprint on the U.S telecommunications market. More concretely, it examines how the merger influenced market share distributions among the four major competitors (AT&T, Verizon, T-Mobile, and Sprint) and the consumer conversion rates between providers (the switching behavior of consumers between providers). By using empirical data, this study aims to provide insights into post-merger market dynamics, helping asses whether the merger led to significant structural changes in competition and consumer behavior.

Motivation

Background and motivation The U.S. telecommunications industry is highly concentrated, with a few dominant firms controlling the market. In 2020, the merger between T-Mobile and Sprint reduced the number of major competitors from four to three, raising concerns about potential market power consolidation, price changes, and consumer choice. The Federal Communications Commission (FCC) and the Department of Justice (DOJ) approved the merger under the assumption that it would enhance competition and benefit consumers. However, its real-world effects remain unstudied.

Understanding the impact of this merger has several implications. First, the merger may have affected conversion rates, affecting consumer retention. Second, fewer competitors in this market may have shifted market dynamics, influencing consumers. Third, regulators need empirical evidence to asses whether the approval of the merger led to the intended outcomes. Lastly, insights from this study can help telecom companies and future entrants make data-driven strategic decisions.

Research Question

How did the 2020 merger between T-Mobile and Sprint impact market share distribution and consumer conversion rates in the U.S. telecommunications market?

Data

The data comes from three data sources.

1. Parquet files: including T-Mobile, Sprint, AT&T, and Verizon user data per CBSA (Core-Based Statistical Areas). This data was obtained from other researchers involved in the project.

Number of Observations: TBD

table of variable description/operstionalisation: TBD

Variable	Description

2. Form 477 data: technology data from the Federal Communications Commission that shows the total coverage per technology per CBSA per merchant. This is public data available online: https://www.fcc.gov/general/broadband-deployment-data-fcc-form-477.

Number of Observations: TBD

table of variable description/operstionalisation: TBD

Variable	Description

3. Safegraph data: this is retail store data that shows the number of stores for a given brand, along with relevant metadata such as where the store is located, when it opened, the number of stores available. This data was obtained with the collaborators working on this project.

Number of Observations: TBD

table of variable description/operstionalisation: TBD

Variable	Description

Method

To analyze the impact of the T-Mobile and Sprint merger on the U.S. telecommunications market, I will use a Fixed Effects (FE) regression model. This approach allows us to control for unobserved heterogeneity across different geographic regions and time periods.

The primary regression model is specified as follows:

MarketShare_{i, c, t} = β_0 + β_1 MergerDate_t + Γ TechnologyControls_{i, t} + δ_c + λ_t + ε_{i, c, t}

And

ConsumerConversionRatio_{j, c, t} = β_0 + β_1 MergerDate_t + Γ TechnologyControls_{j, t} + δ_c + λ_t + ε_{j, c, t}

Where:

• MarketShare_{i, c, t} = Market share of company i in CBSA c at time t.
• ConsumerConversionRatio_{i, c, t} = Loss conversion ratio i in CBSA c at time t.
• MergerDate_t = Indicator variable that equals 1 after the T-Mobile/Sprint merger date and 0 otherwise.
• TechnologyControls_{i, t} = A set of controls accounting for differences in network technologies (e.g., 4G, 5G deployment).
• δ_c = CBSA Fixed Effects to control for region-specific factors.
• λ_t = Monthly Fixed Effects to account for time trends and seasonality.
• ε_{i, c, t} = The error term capturing unexplained variation.

Why this model? I choose the Fixed Effects (FE) approach for the following reasons:

1. Controls for Regional Differences: Different CBSAs may have unique market conditions, and including CBSA Fixed Effects (δ_c) helps eliminate bias from these differences.
2. Accounts for Time Trends: Monthly Fixed Effects (λ_t) control for external shocks, such as economic conditions or industry-wide changes in pricing strategies.
3. Distinguishes Causal Impact of the Merger: The inclusion of a Merger Date dummy ((MergerDate_t) allows to assess whether market share significantly changed after the merger.
4. Technology Controls: Network improvements, such as the rollout of 5G, may independently affect market share, so I include technology variables to avoid confounding effects.

Preview of Findings

• Describe the gist of your findings (save the details for the final paper!)
• How are the findings/end product of the project deployed?
• Explain the relevance of these findings/product.

Repository Overview

*Include a tree diagram that illustrates the repository structure

Dependencies

Explain any tools or packages that need to be installed to run this workflow.

Running Instructions

Provide step-by-step instructions that have to be followed to run this workflow.

About

This project is set up as part of the Master's course Data Preparation & Workflow Management at the Department of Marketing, Tilburg University, the Netherlands.

The project is implemented by team < x > members: < insert member details>