4 Empirical Approach and Results
4.1 Struggling to Keep Up
As discussed in Section 3, the financial distress measure is constructed using responses to the
question on how well respondents are keeping up with their bills and credit commitments.
In particular, I create a dummy variable “Struggle to Keep Up” which is equal to one
for all respondents reporting some degree of struggle in keeping up with their bills and
credit commitments, and equal to zero for those who report no difficulties at all in keeping
up with bills and credit commitments. I exclude respondents who have no bills / credit
commitments or who either refused to answer the question or reported that they did not
know the answer, though as shown in Table 1, these categories represent less than 1 per
cent of the total sample.
Since the dependent variable “Struggle to Keep Up” is a binary variable, I use discrete
dependent variable techniques to examine the impact of the various demographic, socio-
economic and behavioural variables on the probability of experiencing financial distress.
Specifically, I specify the following probit model:
Prob (yi =1) = F (βxi) + ∈i i = 1,2, ...n
where y is the dependent variable “Financial Distress”, x comprises a set of characteris-
tics posited to influence the presence of financial distress (including demographic, socio-
economic and behavioural variables), β is a set of parameters to be estimated, e is the error
term and i is the observation number.
In Table 6, I describe the various independent variables that are used in the analysis.
The probit results are presented in Table 7, where the estimated marginal effects and
standard errors of the parameters for the probit regressions are reported. These marginal
effects are calculated at the means of the independent variables. The likelihood ratio (LR)
test results and the McFadden R2 are also shown.
Demographics and Income: I begin by examining the role of demographics and income
in the ability of respondents to keep up with their bills and credit commitments. These
results are shown in the first column of Table 7. I find, as expected, that marital status,
the number of dependent children, age, unemployment, education and income all matter
for financial distress.
15