Back to Top

 Skip navigation

Limitations of Results and Future Plans

CSO Frontier Series Research Paper

CSO statistical publication, , 11am
Frontier Series Output

CSO Frontier Series outputs may use new methods which are under development and/or data sources which may be incomplete, for example new administrative data sources. Particular care must be taken when interpreting the statistics in this release.
Learn more about CSO Frontier Series outputs

Limitations of Results

Household Budget Survey (HBS)

The HBS data used to produce this household characteristics analysis of the CPI is from 2015/16. The survey usually takes place every five years but was delayed because of the COVID-19 pandemic. Collection of data for the HBS started in July 2022 and will continue for at least 12 months. The HBS data will then need to be analysed before being used to update the CPI basket of goods and services and rebase the CPI.

The years since the 2015/16 HBS have seen considerable changes, notably the disruptions to society and economy caused by the pandemic particularly over the course of 2020/21, resultant changes in consumption behaviour, international supply chain problems, energy-related price inflation, continuing pressures in the housing market, climate change, and the more recent impacts of the war in Ukraine. While the HBS will not distinguish any of these factors individually, we can reasonably expect that the consumption patterns measured by the 2022/23 HBS will be different from 2015/16. Some goods and services will account for a greater share of private household expenditure, and some a lower share, in 2022/23 than in 2015/16. It can also be expected that differences between categories of households may have changed. These changes will, in turn, impact on the calculation of estimates of inflation by household characteristics. The extent of these changes will not be known until results are available from the 2022/23 HBS.

Pricing assumptions inherent in CPI compilation

In compiling a breakdown of the CPI by household groups, an inherent assumption is that all households and household groups are subject to the same price changes. This may not be true. In the overall CPI, this issue is addressed by selecting a broad range of price quotations for each item. However, households in lower income deciles may tend to buy different brands and shop at different outlets than households in higher income deciles. The price changes for these different brands and outlets may differ. Another example is that rural households may buy more items from smaller shops and less from national chains than urban households, and again the price changes may differ. In estimating a breakdown of inflation by household categories, this analysis aims to take account of the different mix of goods and services purchased by each category of household. However, it cannot take into account differences in the prices – and crucially the monthly changes in those prices – encountered by each category of household.

Understanding results that seem counter-intuitive

The results can sometimes appear counter-intuitive. For example, Table 3.4 shows that households paying a mortgage had lower inflation than households that own their homes outright. While households paying a mortgage probably do have higher outgoings than a household that owns their home outright, mortgage interest payments had relatively low inflation over the period covered by this analysis. Thus, households making large mortgage interest payments would have seen overall lower inflation.

Another counter intuitive result from Table 3.13 is that two adults with children have lower inflation than households with just two adults. Households with children have higher proportions of spending on food and clothing. These had relatively low inflation over this period, so the inflation for households of two adults with children is lower.

Even when the results seem intuitive, it is important to realise that they are not inevitable. For example, Table 3.1 shows that households in lower income deciles had higher estimated inflation than households in higher income deciles. This is because households with lower incomes spend a higher proportion of their expenditure on Rent and Electricity, Gas & Other Fuels, and these items had high inflation over the period. But households with higher incomes spend a higher proportion of their income on Transport and on Restaurants & Hotels, and price increases in these sectors may, at certain times, drive their estimated inflation higher than households with lower incomes.

Future Plans

This research paper is the third report from the CSO giving an estimated breakdown of the Consumer Price Index by household characteristics. The results continue to be presented as a CSO Frontier Series, meaning that the methods used and the conclusions from the analysis are subject to further review and refinement.

This analysis is an update of the CSO’s first two reports on this topic: Estimated Inflation by Household Characteristics, March 2022 and Estimated Inflation by Household Characteristics, June 2022. These papers were informed by the work published earlier in 2022 by the Central Bank of Ireland on Household characteristics, Irish inflation and the cost of living (see the Background section of the Introduction chapter) and by similar analyses published by other National Statistical Institutes. While there is no international standard practice as of now in relation to methodology or publication of consumer price indices broken down by household characteristics, the CSO will continue to monitor international best practice in this field.

The CSO will continue to liaise with key users and monitor the level of interest in this topic and in further analysis, and any inputs from users regarding the methodology. As noted in the Background section of the Introduction chapter, the annual rate of inflation was below 2% from September 2012 to June 2021 inclusive. While prices decreased on an annual basis through much of 2020, inflation has been increasing since April 2021. In July 2021, annual inflation was 2.2% and the rate of inflation continued to rise for the rest of the year, exceeding 5% in each of the final three months of 2021. Inflation so far in 2022 has been 5.0% in January, 5.6% in February, 6.7% in March, 7.0% in April, 7.8% in May, 9.1% in both June and July, 8.7% in August, 8.2% in September, and 9.2% in October.

This increased rate of inflation has prompted greater interest in the differing impacts of inflation on different types of household. During the years when annual inflation was below 2%, this analysis would only have shown small differences by type of household.

The rate of inflation in the coming months and years will therefore be a key factor in determining how often this analysis will be repeated but, for now, the CSO plans to repeat this analysis and publish the results from it on a quarterly basis. Thus, the current expectation is that CSO will publish a December 2022 update to this analysis in early 2023.

The results from the 2022/23 HBS will also be a key benchmark for repeating this analysis, using the most recent data on household consumption patterns. Collection of the HBS began in early July 2022 and will continue for at least 12 months. The HBS data will then need to be analysed before being used for the next rebase of the CPI. After that, a report could be prepared using the 2022/23 HBS data to estimate the breakdown of the CPI by household characteristics.