Retail Trade Survey Research Articles

Bayesian Multisource Hierarchical Models with Applications to the Monthly Retail Trade Survey

Abstract The integration of multiple survey, administrative, and third-party data offers the opportunity to innovate and improve survey estimation via statistical modeling. With decreasing response rates and increasing interest for more timely and geographically detailed estimates, imputation methodology that combines multiple data sources to adjust for low unit response and allow for more detailed publication levels, including geographic estimates, is both timely and necessary. Motivated by the Advance Monthly Retail Trade Survey (MARTS) and Monthly Retail Trade Survey (MRTS), we propose Bayesian hierarchical multiple imputation-dependent data models with the goals of automating imputation for the MARTS by using historic MRTS data and providing geographically granular (state-level) estimates for the MRTS via mass imputation using third-party data and spatial dependence. As a natural byproduct of this approach, measures of uncertainty are provided. This article illustrates the advantages of applying established Bayesian hierarchical modeling techniques with multiple source data to address practical problems in official statistics and is, therefore, of independent interest. The motivating empirical studies are unified by their hierarchical modeling framework, which ultimately results in a more principled approach for estimation for the MARTS and a more geographically granular data product for the MRTS.

The concerning increasing trend of alcohol beverage sales in the U.S. during the COVID-19 pandemic.

There are some concerns regarding alcohol use behaviors during the COVID-19 pandemic. The mixed findings of the first alcohol use studies during this pandemic may reflect the lack of differentiation between on-premise and home consumption. Most of the countries adopted severe restrictions on drinking place functioning. Alcohol retail store sales temporal data were used to examine alcohol sales changes in the United States (U.S.) throughout the COVID-19 pandemic as a proxy indicator of at-home drinking. Data were sourced from the Monthly Retail Trade Survey, which has provided U.S. representative estimates of sales at retail and food services stores since 1951. In the present study, we analyzed data from seasonally adjusted beer, wine, and liquor store (BWLS) sales from January 1992 to September 2020. Poisson cubic spline models were used to assess nonlinearity in such sales during the period. These models were adjusted to the consumer price index for alcoholic beverages. There was a significant increase in retail alcohol sales during the beginning of the pandemic, reaching a plateau in the third quarter of 2020. During the COVID-19 period (March 2020 to September 2020), there were 41.9 billion dollars in BWLS sales, representing an increase of 20% compared to the same period in 2019. On the other hand, food and drinking place retail sales decreased by 27% during the same period in the same survey. These results may indicate an increase in home drinking during the period, which could potentially lead to higher alcohol consumption and alcohol-related adverse health outcomes. More aggressive efforts should be made to warn the population about the risks associated with increased home alcohol consumption during the pandemic. Additionally, tracking individual alcohol consumption and releasing real-time data at different levels are needed to better assess the effects of increased alcohol consumption during the pandemic.

Evolution of Business and Consumer Uncertainty in the Midst of the COVID-19 Pandemic. a Sector Analysis in 32 European Countries

This paper examines the evolution of business and consumer uncertainty amid the Coronavirus pandemic in 32 European countries and the European Union (EU). Since uncertainty is not directly observable, we approximate it using the geometric discrepancy indicator of Claveria et al. (2019). This approach allows us quantifying the proportion of disagreement in business and consumer expectations of 32 countries. We have used information from all monthly forward-looking questions contained in Joint Harmonised Programme of Business and Consumer Surveys conducted by the European Commission (EC): the industry survey, the service survey, the retail trade survey, the building survey and the consumer survey. First, we have calculated a discrepancy indicator for each of the 17 survey questions analysed, which allows us to approximate the proportion of uncertainty about different aspects of economic activity, both form the demand and the supply sides of the economy. We then use these indicators to calculate disagreement indices at the sector level. We graphic the evolution of the degree of uncertainty in the main economic sectors of the analysed economies up to June 2020. We observe marked differences, both across variables, sectors and countries since the inception of the COVID-19 crisis. Finally, by adding the sectoral indicators, an indicator of business uncertainty is calculated and compared with that of consumers. Again, we find substantial differences in the evolution of uncertainty between managers and consumers. This analysis seeks to offer a global overview of the degree of economic uncertainty in the midst of the Coronavirus crisis at the sectoral level.

Using electronic transaction data to add geographic granularity to official estimates of retail sales

IntroductionEconomists are interested in more granular, more frequent data to aid in their understanding of the U.S. economy. The most frequent economic data currently available from the U.S. Census Bureau come from monthly economic indicators such as the Monthly Retail Trade Survey, which produces national estimates of retail sales. On the other hand, the most granular data (in terms of geographic and industry detail) come from the Economic Census, which is conducted every five years. The Census Bureau is researching whether organic, third-party Big Data sources, in conjunction with survey data, allow for the production of retail sales estimates that are both monthly and subnational.Case descriptionThis case study explores the feasibility of using aggregated electronic transaction data from First Data (FD), a large payment processor, to calculate experimental regional and state-level monthly estimates of retail sales. Quality criteria are devised to understand this data source’s representativeness of the target population and consistency with existing survey data. Five retail industries in the FD transaction data are identified as having acceptable quality for estimation. Estimation methodology is developed based on linear mixed models in a Bayesian framework. These models try to take advantage of the timeliness of the FD transaction data and smooth over artifacts of FD’s business activity. Experimental estimates of retail sales are calculated for the period January 2015 through March 2018.Discussion and evaluationThe experimental estimates are evaluated quantitatively via correlations between external estimates of the number of employees by industry and qualitatively with respect to additional information about the economy. Many features of the experimental estimates seem reasonable, but there are also caution flags such as anomalous trends related to identified FD quality issues.ConclusionsThe FD transaction data offer insight into economic activity at a more granular level. However, using this data source to enhance official estimates of retail sales is challenging; the FD aggregates have limitations in terms of suppression, coverage, and trends. Consequently, fewer industries than expected are identified as having acceptable quality for estimation. Future work involves calculating experimental estimates for more recent months and researching alternative methods for evaluating their accuracy.

From Transactions Data to Economic Statistics: Constructing Real-time, High-frequency, Geographic Measures of Consumer Spending

Access to timely information on consumer spending is important to economic policymakers. The Census Bureau's monthly retail trade survey is a primary source for monitoring consumer spending nationally, but it is not well suited to study localized or short-lived economic shocks. Moreover, lags in the publication of the Census estimates and subsequent, sometimes large, revisions diminish its usefulness for real-time analysis. Expanding the Census survey to include higher frequencies and subnational detail would be costly and would add substantially to respondent burden. We take an alternative approach to fill these information gaps. Using anonymized transactions data from a large electronic payments technology company, we create daily estimates of retail spending at detailed geographies. Our daily estimates are available only a few days after the transactions occur, and the historical time series are available from 2010 to the present. When aggregated to the national leve l, the pattern of monthly growth rates is similar to the official Census statistics. We discuss two applications of these new data for economic analysis: First, we describe how our monthly spending estimates are useful for real-time monitoring of aggregate spending, especially during the government shutdown in 2019, when Census data were delayed and concerns about the economy spiked. Second, we show how the geographic detail allowed us quantify in real time the spending effects of Hurricanes Harvey and Irma in 2017.

The rise and rise of dining out in New Zealand

The 2018 Restaurant Association Hospitality Report [1] shows Kiwis are eating out more often, for a wider range of occasions. The latest report finds that nationwide sales for the hospitality industry have continued to grow, with takeaway food recording the highest growth.
 In 2018, New Zealand’s hospitality sector achieved record sales of over $11.2 billion (year end, March). This represents sales growth of 3.6 percent over the previous year, which after two years of significant growth (8.2 percent from 2016–2017 and 9.7 percent from 2015–2016) settles at a more stabilised level in 2018. Conversely, EFTPOS data shows that grocery sales are continuing to slow, pointing to people eating out more often, replacing meals that may traditionally have been eaten at home. Over the past five years, there has been a slowdown in year-on-year supermarket sales growth from 4.9 percent in 2014 to 3.9 percent this year [2].
 A recent My Food Bag and Stuff survey showed that only 52 percent of parents now eat at home every night. Statistics NZ data [3] shows that more than a quarter (26 percent) of all food spending is now at restaurants and on ready-to-eat meals, such as takeaway hot drinks and takeaway pizzas (compared with 23 percent in 2014). The takeaway/food-to-go sector is recording the highest growth. Sales for the food-to-go sector grew 5.7 percent in 2018. In dollar terms, this translates to an increase in annual sales of $148 million. But it seems we Kiwis are still hooked on dining out, with restaurants and cafés the biggest winners and accounting for $5.6 billion of all hospitality sales.
 Consumer spending is highest in Auckland, Wellington and Christchurch. These three regions all have annual sales of more than $1 billion per annum. The Ministry of Business Innovation and Employment (MBIE) forecast an annual employment growth for the hospitality sector of 2.7 percent per annum through to 2026. For the period 2016–2017, however, the industry achieved employment growth of almost three times that, at 6.8 percent. The total number of people employed in hospitality is now just under 130,000, with more than 72,000 in restaurants and cafés.
 Hospitality business owners rank their number one challenge as the lack of skilled employees, followed by managing wage costs. This competition for skilled employees has the potential to drive wage rises in some regions, although operators also look for creative ways to retain employees to ensure their labour costs are kept under control. Wages have the potential to rise beyond customers’ expectation of price rises, and that’s a challenge and a balancing act that hospitality business owners must face. The challenge for hospitality owners to find staff is compounded by the number of new businesses opening every week, although to a certain extent this is offset by a comparable number of businesses closing. In 2017, while more than 2,700 new businesses opened, due to those closures, the volume of new outlets overall was an increase of 534 new establishments.
 The hospitality industry has performed exceptionally well in recent years and, although 2018 sees more restrained growth, the industry is well positioned to face the challenges of its competitive operating environment. Although a more cautious outlook is expected for the remainder of 2018, there are also opportunities for operators – particularly for those that deliver an exceptional offering to customers and for those that embrace both changing consumer dining trends and developments in technology to help grow their businesses.
 Highlights:
 
 Nationwide sales for the hospitality industry in 2018 (year end, March) increased by 3.6 percent, to exceed $11 billion.
 The sales growth in 2018 carried across all sectors, excluding the clubs sector, with takeaway/food-to-go recording the highest growth of 5.7 percent.
 Regionally, revenue growth in the Bay of Plenty region was highest for the second year in a row at 6.8 percent, followed by Auckland at 5.1 percent.
 In 2017, the number of hospitality businesses nationwide increased by 534 to 17,328.
 The industry currently employs almost 130,000 people.
 The top challenges identified by the industry are a lack of skilled employees, wage costs, and building and maintaining sales.
 
 Corresponding author
 Marisa Bidois can be contacted at: marisa@restaurantnz.co.nz
 References
 (1) Restaurant Association. Annual Hospitality Report 2018, 2018. https://www.restaurantnz.co.nz/product/2018-hospitality-report/
 (2) Marketview. Consumer Spending Year Ending June 2018. https://marketview.co.nz/news/ (accessed Aug 18, 2018).
 (3) Stats NZ. Retail Trade Survey: March 2018 Quarter. https://www.stats.govt.nz/information-releases/retail-trade-survey-march-2018-quarter (accessed Sep 5, 2018).

Fractional Imputation in Survey Sampling: A Comparative Review

Fractional imputation (FI) is a relatively new method of imputation for handling item nonresponse in survey sampling. In FI, several imputed values with their fractional weights are created for each missing item. Each fractional weight represents the conditional probability of the imputed value given the observed data, and the parameters in the conditional probabilities are often computed by an iterative method such as EM algorithm. The underlying model for FI can be fully parametric, semiparametric, or nonparametric, depending on plausibility of assumptions and the data structure. In this paper, we give an overview of FI, introduce key ideas and methods to readers who are new to the FI literature, and highlight some new development. We also provide guidance on practical implementation of FI and valid inferential tools after imputation. We demonstrate the empirical performance of FI with respect to multiple imputation using a pseudo finite population generated from a sample in Monthly Retail Trade Survey in US Census Bureau.

Detecting and Treating Verified Influential Values in a Monthly Retail Trade Survey

Abstract In survey data, an observation is considered influential if it is reported correctly and its weighted contribution has an excessive effect on a key estimate, such as an estimate of total or change. In previous research with data from the U.S. Monthly Retail Trade Survey (MRTS), two methods, Clark Winsorization and weighted M-estimation, have shown potential to detect and adjust influential observations. This article discusses results of the application of a simulation methodology that generates realistic population time-series data. The new strategy enables evaluating Clark Winsorization and weighted M-estimation over repeated samples and producing conditional and unconditional performance measures. The analyses consider several scenarios for the occurrence of influential observations in the MRTS and assess the performance of the two methods for estimates of total retail sales and month-to-month change.

Multi-level composite estimators and the evaluation of the three-way balanced multi-level rotation designs

The general concept of the three-way balanced multi-level rotation design has been discussed. It has four elements expressing the number of reports in a interview, the number of months for a unit to be in sample and out of sample, and the number of interview waves. However, the role of each element in the three-way balanced multi-level rotation design remains unevaluated, since past research for rotation designs only focused on the balancing conditions and effects and making efficient estimators under the correlation structure from the rotation designs. In this paper, we evaluate effects of the elements in the three-way balanced multi-level rotation designs using numerical studies. We propose multi-level composite estimators (MCEs) and introduce old and new designs from the US Monthly Retail Trade Survey, as criteria in comparative studies with some alternative designs. We also derive variances and mean squared errors of the MCEs, assuming the correlation from the measurements of the same sample unit and the three types of biases in the monthly sample. MCEs are used in numerical studies as main tools to measure the design efficiency of the three-way balanced multi-level rotation designs with three types of biases and revisions between preliminary estimators and final estimator.

An alternative to regression procedure to produce area sample breakdowns of the monthly national estimates of the retail trade survey

The paper describes an alternative to regression procedure developed by Woodruff (1966) to produce estimates of geographic subsets of the original population from a national sample. The proposed procedure always produces positive estimates consistent with each other and with the regional and national estimates. The procedure, however, will be less efficient when sample unbiased estimates and corresponding census data are highly correlated. The results are illustrated with a numerical example.

Improving the Efficiency of Data Collection: A Generic Respondent Follow-up Strategy for Economic Surveys

A major thrust of some large statistical agencies with regard to their survey operations over the past few years has been to reduce respondent burden, increase the efficiency of their production process, improve the quality of end products, and respond to new information needs. To meet these objectives, a rethinking of some production processes has been necessary. This article describes an overall strategy for respondent follow-up in economic surveys. The results of a trial application of the proposed strategy to the Canadian Annual Retail Trade Survey are presented for illustration purposes.

The effect of sampling error on the time series behavior of consumption data

Much empirical economic research today involves estimation of tightly specified time series models that derive from theoretical optimization problems. Resulting conclusions about underlying theoretical parameters may be sensitive to imperfections in the data. We illustrate this fact by considering sampling error in data from the Census Bureau's Retail Trade Survey. We find that parameter estimates in seasonal time series models for retail sales are sensitive to whether a sampling error component is included in the model. We conclude that sampling error should be taken seriously in attempts to derive economic implications by modeling time series data from repeated surveys.

Composite Estimation in Finite Populations

Abstract It is common practice in large recurring surveys to employ rotation sampling schemes. Composite estimators are often used in association with rotation schemes to reduce the variability of the survey estimators. In this article, expressions for the variances of certain composite estimators are derived. It is demonstrated how the variance expressions may be used in planning panel surveys and in certain time series analyses. This work is illustrated by the Census Bureau's Retail Trade Survey.

The Use of Rotating Samples in the Census Bureau's Monthly Surveys

Abstract Two procedures which can be used to increase the reliability of estimates from rotating samples are described. One is the composite estimation procedure which in effect weights results from past and present rotating panels to produce an efficient estimate for the current period. The other procedure is the large observation procedure in which unusually large observations from two or more panels are enumerated currently with reduced weights. The latter procedure is particularly effective in reducing the variance of estimates of month-to-month change. Applications of the procedures to the Monthly Retail Trade Survey of the Bureau of the Census are described. In this survey use of the procedures has reduced the variances of dollar volume estimates and ratios of these estimates (such as month-to-month or year-to-year ratios) far below the variances obtained from a nonrotating sample. These striking variance reductions are achieved with only a slight increase in cost over a nonrotating sample.