Public perceptions of benefit fraud
Trust in social welfare institutions relies on how the public perceives of the deservingness of recipients of
such benefits. In particular, notions of so-called ‘benefit cheats’ erodes public confidence in social welfare.
Past surveys have shown that the public tend to greatly over-estimate the financial scale of benefit fraud.
For example, a 2013 Ipsos MORI study found that on average, respondents guessed that £24 out of every
£100 in benefit claims is done fraudulently, whereas the official estimate is around £0.70 of every £100.
In this section, we explore public perceptions about fraudulently claiming benefits, and in particular,
about how widespread false benefits claims are. More specifically, we will examine whether certain socio
demographic characteristics are associated with public perceptions about how widely false benefit claims are
made in the UK.
We will use part of the British Social Attitudes (BSA) data set on poverty and wealth, which you can
download as bsa-poverty.csv from the PUBL0055 Moodle page. The data set contains the following
Variable name & Description
Answer to the question “Out of every 100 people receiving benefits in
Britain, how many have broken the law by giving false information to
support their claim?”
Five-point left-right ideological scale, with 0 to the left and 4 to the right
RSex Sex of respondent, 1 if male or 0 if female
Completed university degree, 1 if the respondent completed degree or 0 if
You can load the data set by using the following command:
bsa <- read.csv(“data/bsa-poverty.csv”)
Questions (42 Marks)
1. How many individuals are included in the data set?
2. Is the left-right ideology similar on average for those who did and not answer the item NatFrEst?
3. Calculate the median of the variable NatFrEst. What does this tell us about the distribution of
perceptions about fraudulent benefits claims?
4. Create a histogram for NatFrEst and interpret it. What does this tell us about public perceptions
about fraudulent benefits claims?
5. We are interested in seeing whether there is a relationship between a person’s left-right orientation
and how widespread they think fraudulent benefit claims are. Fit the relevant simple linear regression
model and interpret the substantive significance of the estimated slope coefficient. You do not need to
discuss statistical significance.
6. State a null and an alternative hypothesis for the estimated slope coefficient, decide whether to reject
the null hypothesis, and provide a conclusion.
7. How is your conclusion in Question 6 related to Type I and Type II error?
EasyDue™ 支持PayPal, AliPay, WechatPay, Taobao等各种付款方式!
E-mail: email@example.com 微信:easydue