Emily Ip and Lily Xie

If you’ve passed through New Jersey for an amount of time, chances are you’ve realized there’s been an underlying rivalry between North and South Jersey for as long as Taylor Ham has been around. New Jersey is such a diverse state, where you’ll find people rooting for different sports teams, using different vernacular slang, and coexisting amongst a wide variety of industries, landscapes, cultures, food, and languages, amongst other things . With the number of people who come from all over to populate such a dense amount of space, there’s bound to be disparities in the amount of opportunity and resources available for everyone. With over 9 million people from all types of backgrounds living in the state of New Jersey, how does it all play out?

The issues of higher education, employment rates, and poverty have become hot-button topics in recent presidential debates and campaigns. Unavoidably, for younger generations the need to get a college degree has become the norm and even necessary for a number of reasons, including greater employability after graduation. But in reality, how much does education really guarantee your ability to secure a job upon graduation? How would a lack of college education change your employability and income bracket? Even more interestingly, with the diversity of language-speaking populations in New Jersey alone, how might the language you speak play a role in the future you have? How do these factors affect income outcomes, potentially resulting in socioeconomic disparities and poverty status? These are the questions we seek to explore.

New Jersey

Before we dive into exploring these questions, a map of New Jersey and its counties will help digest the visualizations that follow:

As we mentioned before, New Jersey is an incredibly diverse state, perhaps because it is sandwiched between New York City from the north and Philadelphia from the south. Situated between these two metropolitan cities, it’s no wonder that New Jersey is also very linguistically diverse. That said, our conversation of poverty in terms of income, education level, and employment status will be analyzed through the lens of language. That is, we will observe the poverty factors as they play out for the following language groups:

Speak English only
Speak Spanish
Speak other Indo-European languages
Speak Asian and Pacific Island Languages
Speak other languages

You’ll notice as you look at the graphs that follow that the data is arranged by frequency (in other words, the number of individuals in the specified group) in descending order by county. What’s interesting is that there seems to be a trend in which the top handful of counties in any given graph are consistently and largely occupied by counties from Northern New Jersey. This is an interesting observation that we will not discuss in this conversation about poverty status in New Jersey, but may be worthy of keeping in mind for future considerations. We include a map of population density from the 2000 U.S Census Bureau data for your reference and consideration while analyzing the following graphs.

Income Status

Using data from the 2017 American Community Survey and the tidyverse & tidycensus packages from R for data visualization purposes, we use these tools to graph poverty status data as the baseline and reference of poverty status in New Jersey by county. We will use this to draw comparisons from other factors, such as education and employment status, to see if we see similar mappings of the data from those variables.

njincome <- get_acs(geography = "county", 
              variables = c("Speak only English" = "B16009_010", "Speak Spanish" = "B16009_011", "Speak other Indo-European languages" = "B16009_012", "Speak Asian and Pacific Island languages" = "B16009_013", "Speak other languages" = "B16009_014"), 
              state = "NJ",
              year = 2017)

ggplot(njincome, aes(x = estimate, y = reorder(NAME, estimate), color = variable)) + scale_x_log10() + geom_point() + geom_jitter() + labs(title = "Adult income below poverty level by county in NJ",
       subtitle = "2017 American Community Survey",
       y = "",
       x = "Estimated number of individuals by ten-thousands", color = "Language spoken")

What factors contribute to the difference in poverty levels?

Education & Employment

In the next four visualizations, we will observe education levels of New Jersey citizens aged 18 and over, grouped by county, who were in the labor force in 2017. The four education levels we will be observing are as follows:

Completed some high school
High school graduate
Some college/associate’s degree
Bachelor’s degree or higher

Level 1: Completed some high school

njedu1 <- get_acs(geography = "county", 
              variables = c("Speak only English" = "B16010_004", "Speak Spanish" = "B16010_005", 
              "Speak other Indo-European languages" = "B16010_006", "Speak Asian and Pacific Island languages" = "B16010_007",        "Speak other languages" = "B16010_008"),
              state = "NJ", year = 2017)

ggplot(njedu1, aes(x = estimate, y = reorder(NAME, estimate), color = variable)) + scale_x_log10() + geom_point() + geom_jitter() + labs(title = "2017 Adult Educational and Employment Status", subtitle = "Completed some high school & in labor force", y = "county in NJ", x = "Number of individuals by ten-thousands", color = "Language spoken")

For those who have completed some high school, English monolinguals and Spanish speakers comprise the highest number of employed individuals with some high school education. There is also a good amount of variation in the data points, particularly when we observe how each language group compares with each other. While there is overlap consistent between groups, such as “Speak only English” and Speak Spanish" who are the forerunners in the labor force with a partial high school education, generally speaking the amount of individuals in the labor force in this education level seems to correlate with the language that they speak in descending order:

Speak only English
Speak Spanish
Speak other Indo-European languages
Speak Asian and Pacific Island languages
Speak other languages

Level 2: High school graduate

njedu2 <- get_acs(geography = "county", 
              variables = c("Speak only English" = "B16010_017", "Speak Spanish" = "B16010_018", 
              "Speak other Indo-European languages" = "B16010_019", "Speak Asian and Pacific Island languages" = "B16010_020",        "Speak other languages" = "B16010_021"),
              state = "NJ", year = 2017)

ggplot(njedu2, aes(x = estimate, y = reorder(NAME, estimate), color = variable)) + scale_x_log10() + geom_point() + geom_jitter() + labs(title = "2017 Adult Educational and Employment Status", subtitle = "High school graduate & in labor force", y = "county in NJ", x = "Number of individuals by ten-thousands", color = "Language spoken")

Similar to English monolinguals with a partial high school education, those who have completed secondary education number highest in employment attainment. This pattern persists across northern and southern counties, with little variation, and sets English speakers apart from other groups. While Spanish speakers display the second highest employment counts, their frequencies show great variation across counties. Overlap occurs for speakers of Indo-European and Asian and Pacific Island languages. “Other” language speakers have the lowest graduation and employment rates of the five groups.

Level 3: Some college or associate’s degree

njedu3 <- get_acs(geography = "county", 
              variables = c("Speak only English" = "B16010_030", "Speak Spanish" = "B16010_031", 
              "Speak other Indo-European languages" = "B16010_032", "Speak Asian and Pacific Island languages" = "B16010_033",        "Speak other languages" = "B16010_034"),
              state = "NJ", year = 2017)

ggplot(njedu3, aes(x = estimate, y = reorder(NAME, estimate), color = variable)) + scale_x_log10() + geom_point() + geom_jitter() + labs(title = "2017 Adult Educational and Employment Status", subtitle = "Some college or associate's degree & in labor force", y = "county in NJ", x = "Number of individuals by ten-thousands", color = "Language spoken")

As evinced in the high school graduation graph, there is a divide between English monolinguals and Spanish speakers. Of the five groups, the former maintains the highest employment frequency, and one with the smallest range. Thus, for English monolinguals to have achieved some education is for them to have mostly positive prospects for employment. While Spanish speakers bridge the success of the English, Indo-European, and Asian-Pacific Island groups, they mostly share the employment rates of the latter two. Again, speakers of “other” languages have the lowest workforce participation.

Level 4: Bachelor’s degree or higher

njedu4 <- get_acs(geography = "county", 
              variables = c("Speak only English" = "B16010_043", "Speak Spanish" = "B16010_044", 
              "Speak other Indo-European languages" = "B16010_045", "Speak Asian and Pacific Island languages" = "B16010_046",        "Speak other languages" = "B16010_047"),
              state = "NJ", year = 2017)

ggplot(njedu4, aes(x = estimate, y = reorder(NAME, estimate), color = variable)) + scale_x_log10() + geom_point() + geom_jitter() + labs(title = "2017 Adult Educational and Employment Status", subtitle = "Bachelor's degree or higher & in labor force", y = "county in NJ", x = "Number of individuals by ten-thousands", color = "Language spoken")

Having a bachelor’s degree or higher cleaves English monolinguals from the four other language groups. Like their partial-college counterparts, English speakers have high employment rates. And unlike patterns in previous educational levels, speakers of Spanish, Indo-European, and Asian-Pacific Island languages cluster the most tightly. Speakers of other languages are also closer in employment numbers to the others.

Conclusion

In summary, we’ve looked at a triad of factors:

Education attainment – 4 levels
Employment status
Language spoken

As much as education is venerated as a democratizing force, it affords no guarantee of success to those who have achieved it. While the above graphs do not address English fluency levels amongst speakers of other languages, they indicate that mastery of the language generally usurps education as the marker for academic and employment success. English monolinguals are consistently well represented in educational attainment and employment levels, independent of county of residence. In addition, for English monolinguals, the greater the education attained, the greater the chances for finding employment. While those with higher education display higher workforce participation than those with secondary education, the inter-group differences remain small. Even with these slight increases in English speakers’ employability, the lowest employment frequency for those who had not completed high school comes in at 1000K. That the value is already high suggests that lack of education is not as great an impediment as it could be.

While employment levels for all English monolinguals had county-agnostic tendencies, those for “other” speakers ranged across counties. In general, speakers from northern counties had higher labor force participation, and those from southern counties, lower. Despite these ranges, employment levels for speakers of languages other than English and Spanish remained similarly low between high school graduates and non-graduates. These low levels continued for those with associates’ degrees and beyond. And for each of the four educational level groups, there were individuals who hovered near or on the 0 value lines. Thus, it is not the achievement of some education that precludes poverty. Rather, the extent to which one can communicate in English bears more strongly on prospects for adequate standards of living.

Education, Employment and Language in New Jersey