Layoffs Analysis

As the COVID-19 pandemic reshapes industries worldwide, the tech sector faces unprecedented challenges, including economic slowdowns and workforce restructuring. In this post, we embark on an exploratory analysis to understand the trends in tech layoffs since the onset of the pandemic and explore the implications of our findings for various stakeholders.

The specific stakeholders asking the question “What are the trends in tech layoffs since the onset of the COVID-19 pandemic, and how have they evolved?” include not only human resources departments of tech firms but also individuals such as new graduates, unemployed professionals, and job seekers within the tech industry.

The answer to this question will inform several key decisions for these stakeholders:

1. Job Search Strategies: For new graduates and unemployed professionals, understanding trends in tech layoffs provides valuable insights into which sectors or regions may offer more promising job opportunities. This information helps them tailor their job search strategies, target industries with lower layoff rates, and identify areas of growth within the tech sector.

2. Career Decisions: Job seekers rely on insights into tech layoff trends to make informed career decisions, such as whether to pursue opportunities in certain industries or specialize in specific skill sets in demand. Understanding the evolving landscape of tech layoffs guides their career trajectory and helps them navigate the job market effectively.

3. HR Departments of Tech Firms: HR departments within tech firms use insights from tech layoff trends to guide workforce planning, refine recruitment strategies, and enhance talent management practices. This understanding enables HR teams to navigate changes in the job market effectively, ensuring the organization remains resilient and competitive amidst evolving industry dynamics.

The dataset that could provide insights into trends in tech layoffs since the onset of the COVID-19 pandemic is the Layoffs dataset from Kaggle. This dataset tracks layoffs reported by tech companies from March 11, 2020, to the present, gathering information from platforms such as Bloomberg, San Francisco Business Times, TechCrunch, and The New York Times.

This data is relevant to the question as it provides comprehensive information on layoffs within the tech industry during the pandemic. It allows for the analysis of trends over time, geographical distribution of layoffs, impact on different industries, and the scale of layoffs reported by various companies. Understanding these trends is crucial for stakeholders such as HR departments of tech firms, individuals seeking jobs, and new graduates, as it informs decision-making related to workforce planning, and career choices within the tech field.

The exploratory data analysis (EDA) involved several steps to gain insights into the dataset and answer the research question. Initially, we imported the dataset into Python using the Pandas library, which allowed us to manipulate and analyze the data effectively. We performed basic data exploration to understand the structure of the dataset, including the columns and their respective data types. This step helped us identify relevant columns for analysis, such as the date of layoff, country, industry, and the number of employees laid off.

Next, we conducted exploratory data analysis to answer the research question regarding trends in tech layoffs. This involved visualizing the data using various plotting libraries such as Matplotlib. We created visualizations such as bar plots, line plots, and histograms to analyze the distribution of layoffs over time, across different countries and industries, and within specific geographical locations. These visualizations provided insights into the patterns and trends in tech layoffs, allowing us to identify key findings and draw meaningful conclusions.

During the initial stages of my analysis, I ensured the integrity of the dataset by addressing missing values, inconsistencies, and duplicates. I handled missing values by dropping rows where the ‘total_laid_off’ column was null, while also converting the ‘location’ and ‘industry’ columns to lowercase to promote consistency. Additionally, I removed any duplicate entries to maintain data cleanliness. Throughout the process, I encountered a significant challenge with parsing date formats, which I addressed by converting the ‘date’ column to a datetime format. I also remained vigilant for inconsistencies in data recording, taking steps to standardize formats where necessary. By implementing these data-cleaning steps and addressing encountered bugs, I was able to proceed with my exploratory analysis, ensuring that my findings were based on reliable and accurate data.

The bar plot displays the total number of tech layoffs reported over a four-year period, covering 2020 to 2023. Each bar represents a single year, with the height indicating the total layoffs reported. A noticeable fluctuation is observed, with a significant increase in layoffs in 2022 and 2023 compared to 2020 and 2021. The highest number of layoffs occurred in 2023, totaling 262,682. This visualization offers a concise snapshot of the overall trend in tech layoffs during the specified timeframe.

Summary of Total Layoffs Over Four Years:

• Year 2020: Total layoffs = 80,998
• Year 2021: Total layoffs = 15,823
• Year 2022: Total layoffs = 164,969
• Year 2023: Total layoffs = 262,682

The bar plot provides a comparative analysis of tech layoffs across different countries, focusing on the top 10 entries sorted by total layoffs. Each bar represents a country, with the height indicating the total number of layoffs reported. The visualization offers insights into the distribution of layoffs among various countries, highlighting those with the highest number of reported layoffs. From the graph, it’s apparent that the United States has the highest number of layoffs, followed by Sweden, Germany, and the Netherlands. Other countries such as the United Kingdom, India, Canada, Singapore, and China also feature in the top 10 entries, albeit with lower numbers of layoffs. This visualization helps to identify the countries most affected by tech layoffs, providing valuable insights for stakeholders interested in understanding global layoff trends.

Summary of Top 10 Entries Sorted by Total Layoffs:

• Country: United States, Total layoffs: 12,000
• Country: Sweden, Total layoffs: 8,500
• Country: Germany, Total layoffs: 8,000
• Country: Netherlands, Total layoffs: 6,000
• Country: United Kingdom, Total layoffs: 2,500
• Country: India, Total layoffs: 2,500
• Country: Canada, Total layoffs: 2,300
• Country: Singapore, Total layoffs: 2,000
• Country: China, Total layoffs: 1,800

The time series plot depicts the monthly trends in tech layoffs throughout the year 2023. Each data point represents a month, with the total number of layoffs reported plotted on the y-axis. The visualization offers insights into the fluctuation of layoffs over the year, highlighting any notable peaks or troughs in activity. From the graph, it’s evident that January saw the highest number of layoffs, followed by February and March. Subsequently, there is a downward trend in layoffs from April to September, with a slight increase observed in October and November before a slight decrease in December. This visualization provides stakeholders with a clear understanding of the monthly dynamics of tech layoffs in 2023, facilitating informed decision-making related to workforce planning and resource allocation.

Summary of Monthly Layoffs Trends in 2023:

• January 2023: Total layoffs = 89,809
• February 2023: Total layoffs = 39,902
• March 2023: Total layoffs = 37,823
• April 2023: Total layoffs = 20,100
• May 2023: Total layoffs = 15,114
• June 2023: Total layoffs = 11,206
• July 2023: Total layoffs = 10,650
• August 2023: Total layoffs = 10,178
• September 2023: Total layoffs = 4,707
• October 2023: Total layoffs = 7,896
• November 2023: Total layoffs = 8,138
• December 2023: Total layoffs = 7,159

While our analysis provides valuable insights, it’s important to recognize its limitations. Firstly, the dataset only covers layoffs up to 2023, which means we’re missing out on any trends or events occurring beyond that timeframe. For instance, we’ve already observed job losses in the early months of 2024, indicating a gap in our understanding of current workforce dynamics. Additionally, the dataset may not capture real-time fluctuations, potentially skewing our perceptions. Moreover, it’s possible that the dataset only includes reported layoffs, leaving out undocumented or unreported job losses. This could particularly affect regions with lax reporting standards or underrepresented industries. Considering recent developments, such as the job losses in early 2024, it appears that the situation may worsen, underscoring the urgency for more comprehensive and up-to-date data analysis to inform decision-making effectively.

GitHub Repository :https://github.com/arshuls/INST414.git