ACS

Intro

I wanted to take a stab at learning Pandas in Python. I am confident that there are plenty of tutorials out there, but instead of trying to find one, I went and found a large dataset, and just decided to google every little thing I wanted to do.

That dataset was the American Community Survey (1-Year) data for 2021. I wanted to see how demographics data differ between the blind and sighted population in the US. Granted, ACS doesn't ask if you're "blind", they ask if you have a vision-related disability. This could mean that you are legally blind without glasses, but can drive with glasses. This could mean you're fully blind without light perception. The ACS doesn't distinguish. Also, it means that the person with visual impairments could have been born blind, or became blind in their 70's. Again, the ACS doesn't distinguish. With those caveats in mind, let's poke at 2GB of data and see what happens.

Counts

How many respondents are we looking at?

Respondents	Number	Percentage
Total	3,252,599	100%
With vision impairments	89,937	2%
Without vision impairments	3,162,662	98%

Demographics

What is the age distribution of our respondents?

How many other disabilities do our respondents have?

There's a lot of overlap between vision difficulties and other difficulties. But, we saw that the distribution of blind people skews much older than the sighted population. What happens if we look at disabilities again, but for a younger population?

Respondents under 50	Number	Percentage
Total	1,906,095	100%
With vision impairments	25,845	1%
Without vision impairments	1,880,250	99%

Education

I wanted to see if there was any difference in degree fields. Now, if I say, "What degrees to blind people have?", that could imply 1 of 2 things: "When a blind person goes for a degree, what do they get?" or "What kinds of degrees cause people to become blind?" However, I don't have access to when someone became blind or when they got their degree, so I figured that the degree field distribution would wind up being pretty muddy.

And then I looked at what the top 5 degree fields were for blind/sighted people.

Blind people's degree field

1. General Education
2. Business Management And Administration
3. Nursing
4. General Business
5. Psychology

Sighted people's degree field

1. Business Management And Administration
2. Psychology
3. General Business
4. Nursing
5. General Education

So, here are those rankings, with numbers attached.

Note that the number of blind respondents with bachelor's or higher is 15,507. For sighted respondents, that number is 852,116.

Field	Blind Rank	Sighted Rank	# (Blind)	% (Blind)	# (Sighted)	% (Sighted)
General Education	1	5	1,165	7.5%	36,059	4.2%
Business Management And Administration	2	1	872	5.6%	48,178	5.6%
Nursing	3	4	779	5%	35,898	4.2%
General Business	4	3	749	4.8%	36,059	4.2%
Psychology	5	2	687	4.4%	38,891	4.5%

Employment

Do people have jobs?

Who do they work for?

For people employed, what industries do they work for?

How do people get to work?

How long does it take people to get to work?

I like looking at the groupings. Sure, some people said it took 47 minutes to get work, but most people just round to 45 or 50.

Speaking of work, how much are people actually making...

Income

Let's hone in on the lower part of that graph. How much money are people making compared to the poverty line?

Let's look at different forms of income for everyone.

For example, 49% of blind respondents have some form of social security income, and 12% have Supplementary Security Income. Compare this to 24% and 3% of sighted respondents, respectively.

Here's the percent of respondents have have some form of different classifications of income:

Here's how social security income compares between populations (for those who get it):

Now, we saw earlier that the majority of people who have vision impairments are above retirement age, which may impact why this population draws disproportionately more social security. So, let's look at those same graphs again, but filtering only from the population below 65.