Digital Divide In Maine

Geographic Inequality in Access to Technology and Digital Resources: An Analysis of Maine's Digital Divide 

Hypothesis: 

Null hypothesis: The Maine Connective Authority (MCA) should prioritize expanding infrastructure beyond Southern Maine, as large parts of Maine experience no or slow internet speed, rather than concentrating on Southern Maine. 


Alternative hypothesis: The Maine Connective Authority (MCA) should concentrate on advancing infrastructure in Southern Maine, despite the fact that large parts of Maine face no or slow internet speed.

Abstract:

The increasing importance of technology and digital resources in today's world has magnified the impact of the digital divide on socioeconomic outcomes. This study aims to measure the inequality in access to technology and digital resources between different geographic areas in Maine. We will use the Digital Divide Index (DDI) to further investigate areas facing socioeconomic challenges and limited broadband access as they will have lower relative scores. Specific factors such as poverty, computers, age, education and broadband download/upload speeds will be analyzed to understand their effects on technological adoption rates. The findings of this study will provide valuable insights for policymakers and stakeholders to make informed decisions about prioritizing infrastructure investments and mitigating the digital divide in Maine.



Background: 

The DDI is  referring to the gap in access to technology and digital resources. It has become a pressing issue in contemporary society. Unequal access to these resources can exacerbate existing social, economic, and educational disparities. We initially were able to see a visual of The Federal Communications Commission (FCC) map which reveals that a majority of Maine experiences no or slow internet speeds, indicating a pronounced digital divide within the state. This divide may negatively impact the social and economic development of affected regions, particularly in rural areas.  The following map shows maximum advertised download speeds across Maine. 



Given these circumstances, it is crucial to examine the factors contributing to this divide and identify regions where efforts to improve broadband infrastructure should be prioritized. In this study, we will explore the relationship between regional differences, socioeconomic challenges, and broadband availability with Digital Divide Index (DDI) scores. By analyzing the association between these factors and DDI scores, we aim to provide a better understanding of the dynamics that underlie the digital divide in Maine. 


A comprehensive analysis of the digital divide in Maine will not only highlight the areas that require urgent attention but also inform policymakers and stakeholders about the most effective strategies to bridge this gap. Focusing on improving broadband infrastructure in the most affected regions could help to level the playing field for communities, fostering equal access to technology and digital resources, and promoting social and economic development across the state



Research Questions: 

Where are the underserved areas in broadband in Maine?

What geographic locations culturally ready for increased infrastructure?

What counties should the MCA prioritize for this years budget? 

Methods:

Methods: To investigate the digital divide in Maine, we used data from the Federal Communications Commission (FCC) and the United States Census Bureau. Our data sources include the FCC broadband data and the Census American Community Survey (ACS) 5-Year Estimate for 2021 . For mathematical purposes, we standardized all the results. By calculating Z-scores for all of the variables we got a standardized representation of data, revealing the position of a specific observation relative to the sample's mean and standard deviation. Because of this, it is important to note that all scores are relative to the state of Maine and the level at which the data has been collected or aggregated. Also note that the DDI score was built to give social economic factors more influence on score. All scores were normalized to fall between a 0 to 100 range, where the higher the number, the higher the digital divide.


The Digital Divide Index (DDI) was calculated based on the following equations:

Equation 1: INFA = NBBND0.3 + NIA0.3 + NCD0.3 – DNS0.05 – UPS*0.05

Equation 2: SE = AGE65 + POV + LTHS + DIS

Equation 3: DDI = INFA + SE


Other equations: 

Equation 4: Culture or Technology = INFA - SE

Equation 5: DDI RATIO = (INFA - SE) / (DDI *not scaled)


DDI Acronyms:

INFA: Infrastructure/adoption score

SE: socioeconomic

NBBND: Percent of population not using internet at 100/20 Mbps

NIA: Percent of population with no internet Access

NCD: Percent of population with no Computing Devices

DNS: Average Download Speed for area

UPS: Average Upload Speed for area

POV: Percent of population at or below poverty level

LTHS: Percent of population over 25 with no High School degree

DIS: Percent of population with disability


Utilizing these five equations and data sources, we conducted a comprehensive analysis of the relationship between regional differences, socioeconomic challenges, and broadband availability with DDI scores. The results of this analysis will help identify the factors contributing to the digital divide in Maine and the regions where efforts to improve broadband infrastructure should be prioritized.


It should be mentioned that the DDI does not account for two key variables: broadband cost and technology usage, as nationwide data is unavailable. Additionally, cellular wireless access is excluded since mobile device constraints and limited data plans diminish the benefits of digital applications and can hinder usage or incur substantial costs.



Preliminary results:

Census Data Correlations:


In Maine, there exists a significant correlation between lack of education, the absence of computing devices in households, poverty, and disability.


The data shows that disability is positively correlated with both lack of education and poverty, suggesting that individuals with disabilities often face greater challenges in accessing quality education and economic opportunities. Additionally, a negative correlation is observed between households without computing devices and factors such as low educational attainment and disability. This indicates that the absence of technology may further exacerbate educational disparities and impede social mobility for individuals with disabilities. 


Upon further investigation, a negative correlation is present between not possessing a computer and poverty, highlighting the importance of technological accessibility in breaking the cycle of poverty. This interconnected web of factors demonstrates the need for targeted policy interventions to address these inequalities and improve the lives of Maine's most vulnerable populations.

Removing FFC Outliers:


The fastest commercially available internet speeds are typically offered by fiber-optic providers, with some offering gigabit connections (1 Gbps or 1000 Mbps) or even faster. For instance, Google Fiber provides speeds of up to 2000 Mbps (2 Gbps) for both residential and small business customers. In some regions, local internet service providers may offer even faster speeds, such as EPB in Chattanooga, Tennessee, which offers a 10 Gbps service. These services are not available in Maine. From our research, the fastest speed in Maine could possibly be near 4,700 from AT&T, however it would have many geographic limitations. Because of this, we truncated internet speed information over 4,700 to 4,700.


This had little to no change on overall results.


Suspected that the data was entered in correctly and there was no way to know what factors they were off by. 


Statiscal summary results showed that the (variables --) largest outliers from this

Evaluating percent of population at or below poverty level:


Poverty in Maine exhibits a distinct spatial pattern, with lower poverty rates observed in the southern portion of the state, while the poverty rate increases as one travels north. In certain areas, the poverty rate reaches at least 17%, which translates to nearly one in five individuals living in poverty. This high poverty rate in large geographic swaths of Maine may be indicative of systemic or regional challenges that need to be addressed.


Further research is needed to uncover the systemic and regional challenges contributing to these disparities, and to develop effective strategies for reducing poverty and promoting equitable opportunities for all residents of the state.


Our analysis shows that numerous census tracts in Maine have poverty rates of 10-20%, indicating moderate poverty levels. Additionally, a few tracts have extreme poverty rates of 40-50%, highlighting areas of high need. A considerable spread in poverty rates (20-30%) in the third quartile reveals substantial disparities within the state and suggests underlying contributing factors. 


Poverty is a major issue as it negatively impacts individuals and communities in various ways, including limited access to quality education, healthcare, and essential resources. It perpetuates social and economic inequalities, hinders human potential, and can lead to increased crime rates, homelessness, and overall reduced quality of life. Addressing poverty is crucial for fostering social cohesion, sustainable development, and ensuring equitable opportunities for all members of society.

Evaluating percent of population over 25 with no High School degree:


Maine exhibits significant contrasts in educational attainment levels between northern and southern regions. Northern Maine has higher rates of individuals without high school diplomas, while southern Maine generally boasts higher rates of educational achievement. However, pockets within southern Maine also report low educational attainment levels, indicating regional discrepancies. 


Various factors such as economic conditions, access to quality education, and resource availability contribute to this disparity. Identifying and addressing these factors is essential to promote equal access to education and ensure all Maine residents have the opportunity to achieve their full potential.



Evaluating percent of population with no Computing Devices:


Maine's digital divide is evident between rural Northern and Western regions and urbanized Southern Maine. Rural areas have higher levels of households without computing devices due to factors like lower incomes, inadequate infrastructure, and limited affordable internet services.


Southern Maine enjoys higher device adoption, benefiting from higher household incomes, better infrastructure, diverse economy, and greater access to affordable internet services. To bridge the digital divide, investments in infrastructure, internet services, and digital literacy programs are needed in underserved areas.



Identifying areas lacking digital literacy or in need of Infrastructure:


The divide may be attributed to cultural or technological factors. We can determine which areas of Maine are divided by either cultural issues or technology based issues by finding the difference between the INFA and SE score. 


If a specific county or census tract has a higher INFA score compared to its SE score, it is essential to prioritize enhancing broadband infrastructure. Conversely, if a certain area has a higher SE score than its INFA score, efforts should be directed towards promoting digital literacy and raising awareness of the benefits of technology.


We find that Maine has very few areas with SE scores higher than INFA scores, so efforts to increase board band are needed almost everywhere. 

Darker blue regions can easily be identified in Southern Maine, especially the Portland metropolitan area. This area serves as the state's economic center, encompassing diverse industries like healthcare, finance, and tourism. This area also boasts a vibrant entrepreneurial environment, with startups and small businesses fueling the economy. 


Broadband infrastructure is vital for businesses to flourish in today's digital landscape. A larger negative score between INFA and SE in Maine signifies a more urgent need to enhance broadband access in the southern region to foster economic growth. 


Investing in broadband infrastructure in Southern Maine will continue to draw new businesses, stimulate innovation, and generate job opportunities, ultimately contributing to the state's overall economic welfare.

Identifying areas culturally ready for improved infrastructure:



To confirm our hypothesis about the importance of investing in areas with a large negative difference between INFA and SE, we developed an equation that takes into account the difference between the two metrics and divides it by the Digital Divide Index (DDI). 


Equation 5: DDI RATIO = ( | INFA - SE | ) / (DDI *not scaled)


By applying this equation, we were able to identify the regions where development should take place to maximize the impact of the MCA’s investments. This approach provides a data-driven solution to address the disparities in broadband access and support economic growth in Maine. See the darker blue areas on the map.


From this visual, it can bee seen that the MCA should focus on Southern Maine. Thus, we reject our null hypothesis.

Final Results and Next Steps for MCA:

We are rejecting the null hypothesis as there is enough evidence to suggest that the alternative hypothesis is true. Therefore, accepting the alternative hypothesis is the logical conclusion.


The Maine Connective Authority (MCA) should concentrate on advancing infrastructure in Southern Maine, despite the fact that the majority of Maine faces no or slow internet speed.


Here we observer the digital divide at the county level similar to the Perdue DDI. 


We have determined that concentrating on Kennebec county, Androscoggin county and Lincoln county could significantly influence Maine's overall economic growth and prosperity. 


These counties exhibit higher INFA - SE scores compared to counties with similar DDI scores, indicating they are culturally prepared for infrastructure improvements. Thus, if the MCA prioritizes these counties, they can be more confident that the resources and support provided will benefit Maine as a whole.


While we also see northern and western counties exhibit significant divisions, however it has been observed that these regions lack a strong cultural foundation surrounding technology. This deficiency can potentially reduce the impact of investments made in these areas, as the local communities may not be fully prepared to capitalize on technological advancements and opportunities.


So, by enhancing broadband infrastructure in these three southern counties, local businesses can be supported, new investments attracted, and job opportunities created for residents. This will stimulate economic activity, raise tax revenues, and contribute to Maine's long-term economic sustainability.

Citations:


Purdue University Research:

"Redlining in America: Systemic Racism and Housing Segregation." Story Maps, Esri, 2021, 

https://storymaps.arcgis.com/stories/8ad45c48ba5c43d8ad36240ff0ea0dc7.


Census Data:

 "American Community Survey." United States Census Bureau, 2021, 

https://api.census.gov/data/2021/acs/acs5/groups/B01001.htm.


FCC data: 

"FCC National Broadband Map." Federal Communications Commission, 2023, 

https://broadbandmap.fcc.gov/home.


Python Language: 

Python Software Foundation. Python Language Reference, version 3.9, 2020, 

https://docs.python.org/3/reference/. 


Geopandas package: 

GeoPandas development team. GeoPandas, version 0.9.0, 2021, https://geopandas.org/. 


Mapclassify package: 

Pyle, David. "mapclassify: Classification Schemes for Choropleth Maps." Journal of Open 

Source Software, vol. 3, no. 29, 2018, https://doi.org/10.21105/joss.00603. 


Requests package: 

Kennedy, Kenneth Reitz. "Requests: HTTP for Humans." Requests, version 2.26.0, 2021, 

https://requests.readthedocs.io/en/master/. 


Numpy package: 

Harris, Charles R., et al. "Array Programming with NumPy." Nature, vol. 585, no. 7825, 2020, 

pp. 357-362, https://doi.org/10.1038/s41586-020-2649-2. 


Seaborn package: 

Waskom, Michael. "Seaborn: Statistical Data Visualization." Seaborn, version 0.11.2, 2021, 

https://seaborn.pydata.org/. 


Matplotlib.pyplot package: 

Hunter, John D. "Matplotlib: A 2D Graphics Environment." Computing in Science & 

Engineering, vol. 9, no. 3, 2007, pp. 90-95, https://doi.org/10.1109/MCSE.2007.55. 


Mapclassify package: 

Pyle, David. "mapclassify: Classification Schemes for Choropleth Maps." Journal of Open 

Source Software, vol. 3, no. 29, 2018, https://doi.org/10.21105/joss.00603.