This data collection contains information gathered in the Survey of Income and Education (SIE) conducted in April-July 1976 by the Census Bureau for the United States Department of Health, Education, and Welfare (HEW). Although national estimates of the number of children in poverty were available each year from the Census Bureau's Current Population Survey (CPS), those estimates were not statistically reliable on a state-by-state basis. In enacting the Educational Amendments of 1974, Congress mandated that HEW conduct a survey to obtain reliable state-by-state data on the numbers of school-age children in local areas with family incomes below the federal poverty level. This was the statistic that determined the amount of grant a local educational agency was entitled to under Title 1, Elementary and Secondary Education Act of 1965. (Such funds were distributed by HEW's Office of Education.) The SIE was the survey created to fulfill that mandate. Its questions include those used in the Current Population Survey regarding current employment, past work experience, and income. Additional questions covering school enrollment, disability, health insurance, bilingualism, food stamp recipiency, assets, and housing costs enabled the study of the poverty concept and of program effectiveness in reaching target groups. Basic household information also was recorded, including tenure of unit (a determination of whether the occupants of the living quarters owned, rented, or occupied the unit without rent), type of unit, household language, and for each member of the household: age, sex, race, ethnicity, marital history, and education.

Abstract

Household Income and Expenditure Survey (HIES) collects a wealth of information on HH income and expenditure, such as source of income by industry, HH expenditure on goods and services, and income and expenditure associated with subsistence production and consumption. In addition to this, HIES collects information on sectoral and thematic areas, such as education, health, labour force, primary activities, transport, information and communication, transfers and remittances, food expenditure (as a proxy for HH food consumption and nutrition analysis), and gender.

The Pacific Islands regionally standardized HIES instruments and procedures were adopted by the Government of Tokelau for the 2015/16 Tokelau HIES. These standards were designed to feed high-quality data to HIES data end users for:

1. deriving expenditure weights and other useful data for the revision of the consumer price index;
2. supplementing the data available for use in compiling official estimates of various components in the System of National Accounts;
3. supplementing the data available for production of the balance of payments; and
4. gathering information on poverty lines and the incidence of poverty in Tokelau.

The data allow for the production of useful indicators and information on the sectors covered in the survey, including providing data to inform indicators under the UN Sustainable Development Goals (SDGs). This report, the above listed outputs, and any thematic analyses of HIES data, collectively provide information to assist with social and economic planning and policy formation.

Geographic coverage

National coverage.

Analysis unit

Households and Individuals.

Universe

The universe of the 2015/16 Tokelau Household Income and Expenditure Survey (HIES) is all occupied households (HHs) in Tokelau. HHs are the sampling unit, defined as a group of people (related or not) who pool their money, cook and eat together. It is not the physical structure (dwelling) in which people live. The HH must have been living in Tokelau for a period of six months, or have had the intention to live in Tokelau for a period of twelve months in order to be included in the survey.

Household members covered in the survey include: -usual residents currently living in the HH; -usual residents who are temporarily away (e.g., for work or a holiday); -usual residents who are away for an extended period, but are financially dependent on, or supporting, the HH (e.g., students living in school dormitories outside Tokelau, or a provider working overseas who hasn't formed or joined another HH in the host country) and plan to return; -persons who frequently come and go from the HH, but consider the HH being interviewed as their main place of stay; -any person who lives with the HH and is employed (paid or in-kind) as a domestic worker and who shares accommodation and eats with the host HH; and -visitors currently living with the HH for a period of six months or more.

Kind of data

Sample survey data [ssd]

Sampling procedure

The 2015/16 Tokelau Household Income and Expenditure Survey (HIES) sampling approach was designed to generate reliable results at the national level. That is, the survey was not designed to produce reliable results at any lower level, such as for the three individual atolls. The reason for this is partly budgetary constraint, but also because the HIES will serve its primary objectives with a sample size that will provide reliable national aggregates.

The sampling frame used for the random selection of HHs was from December 2013, i.e. the HH listing updated in the 2013 Population Count.

The 2015/16 Tokelau HIES had a quota of 120 HHs. The sample covered all three populated atolls in Tokelau (Fakaofo, Nukunonu and Atafu) and the sample was evenly allocated between the three atoll clusters (i.e., 40 HHs per atoll surveyed over a ten-month period). The HHs within each cluster were randomly selected using a single-stage selection process.

In addition to the 120 selected HHs, 60 HHs (20 per cluster) were randomly selected as replacement HHs to ensure that the desired sample was met. The replacement HHs were only approached for interview in the case that one of the primarily selected HHs could not be interviewed.

Mode of data collection

Face-to-face [f2f]

Research instrument

The questionnaires for this Household Income and Expenditure Survey (HIES) are composed of a diary and 4 modules published in English and in Tokelauan. All English questionnaires and modules are provided as external resources.

Here is the list of the questionnaires for this 2015-2016 HIES: - Diary: week 1 an 2; - Module 1: Demographic information (Household listing, Demographic profile, Activities, Educational status, Communication status...); - Module 2: Household expenditure (Housing characteristics, Housing tenure expenditure, Utilities and communication, Land and home...etc); - Module 3: Individual expenditure (Education, Health, Clothing, Communication, Luxury items, Alcohonl & tobacco); - Module 4: Household and individual income (Wages and salary, Agricultural and forestry activities, Fishing gathering and hunting activities, livestock and aquaculture activities...etc).

Cleaning operations

All inconsistencies and missing values were corrected using a variety of methods: 1. Manual correction: verified on actual questionnaires (double check on the form, questionnaire notes, local knowledge, manual verifications) 2. Subjective: the answer is obvious and be deducted from other questions 3. Donor hot deck: the value is imputed based on similar characteristics from other HHs or individuals (see example below) 4. Donor median: the missing or outliers were imputed from similar items reported median value 5. Record deletion: the record was filled by mistake and had to be removed.

Several questions used the hotdeck method of imputation to impute missing and outlying values. This method can use one to three dimensions and is dependent on which section and module the question was placed. The process works by placing correct values in a coded matrix. For example in Tokelau the “Drink Alcohol” questions used a three dimension hotdeck to store in-range reported data. The constraining dimensions used are AGE, SEX and RELATIONSHIP questions and act as a key for the hotdeck. On the first pass the valid yes/no responses are place into this 3-dimension hotdeck. On the second pass the data in the matrix is updated one person at a time. If a “Drink Alcohol” question contained a missing response then the person's coded age, sex and relationship key is searched in the “valid” matrix. Once a key is found the result contained in the matrix is imputed for the missing value. The first preferred method to correct missing or outlying data is the manual correction (trying to obtain the real value, it could have been miss-keyed or reported incorrectly). If the manual correction was unsuccessful at correcting the values, a subjective approach was used, the next method would be the hotdeck, then the donor median and the last correction is the record deletion. The survey procedure and enumeration team structure allow for in-round data entry, which gives the field staff the opportunity to correct the data by manual review and by using the entry system-generated error messages. This process was designed to improve data quality. The data entry system used system-controlled entry, interactive coding and validity and consistency checks. Despite the validity and consistency checks put in place, the data still required cleaning. The cleaning was a two-stage process, which included manual cleaning while referencing the questionnaire, whereas the second stage involved computer-assisted code verification and, in some cases, imputation. Once the data were clean, verified and consistent, they were recoded to form a final aggregated database, consisting of: Person level record - characteristics of every (household) HH member, including activity and education profile; HH level record - characteristics of the dwelling and access to services; Final aggregated income - all HH income streams, by category and type; Final aggregated expenditure - all HH expenditure items, by category and type.

The cleaning was a two-stage process, which included manual cleaning while referencing the questionnaire, whereas the second stage involved computer-assisted code verification and, in some cases, imputation. Once the data were clean, verified and consistent, they were recoded to form a final aggregated database.

Response rate

Overall, 99% of the response rate objective was achieved.

Sampling error estimates

Refer to Appendix 2 of the Tokelau 2015/2016 Household Income and Expenditure Survey report attached as an external resource.

This data collection is part of a longitudinal survey designed to provide detailed information on the economic situation of households and persons in the United States. These data examine the distribution of income, wealth, and poverty in American society and gauge the effects of federal and state programs on the well-being of families and individuals. There are three basic elements contained in the survey. The first is a control card that records basic social and demographic characteristics for each person in a household, as well as changes in such characteristics over the course of the interviewing period. These include age, sex, race, ethnic origin, marital status, household relationship, education, and veteran status. Limited data are provided on housing unit characteristics such as units in structure, tenure, access, and complete kitchen facilities. The second element is the core portion of the questionnaire, with questions repeated at each interview on labor force activity, types and amounts of income, and participation in various cash and noncash benefit programs for each month of the four-month reference period. Data for employed persons include number of hours and weeks worked, earnings, and weeks without a job. Nonworkers are classified as unemployed or not in the labor force. In addition to providing income data associated with labor force activity, the core questions cover nearly 50 other types of income. Core data also include postsecondary school attendance, public or private subsidized rental housing, low-income energy assistance, and school breakfast and lunch participation. The third element consists of topical modules, which are a series of supplemental questions asked during selected household visits. Topical modules include some core data to link individuals to the core files. The Wave 1 Topical Module covers recipiency and employment history. The Wave 2 Topical Module includes work disability, education and training, marital, migration, and fertility histories, and household relationships. The Wave 3 Topical Module covers medical expenses and utilization of health care, work-related expenses and child support, assets and liabilities, real estate, shelter costs, dependent care and vehicles, value of business, interest earning accounts, rental properties, stocks and mutual fund shares, mortgages, and other assets. The Wave 4 Topical Module covers disability, taxes, child care, and annual income and retirement accounts. Data in the Wave 5 Topical Module describe child support, school enrollment and financing, support for nonhousehold members, adult and child disability, and employer-provided health benefits. Data in the Wave 6 Topical Module provide information on medical expenses, work-related expenses and child support paid, assets and liabilities, real estate, shelter costs, dependent care and vehicles, value of business, interest-earning accounts, rental properties, stock and mutual fund shares, mortgages, other financial investments. Wave 7 Topical Module includes annual income and retirement accounts, home health care, retirement expectations and pension plan coverage, and taxes. Wave 8 Topical Module covers adult well-being and welfare reform. Wave 9 Topical Module is the same as Waves 3 and 6 Topical Modules. Wave 10 Topical Module focuses on work schedules, disablility, taxes, child care, and annual income and retirement. Wave 11 includes child support, support for nonhousehold members, and adult and child disability. Wave 12 Topical Module is the same as Waves 3, 6, and 9 but also includes child well-being. (Source: ICPSR, retrieved 06/28/2011)

Long-term longitudinal dataset with information on generational links and socioeconomic and health conditions of individuals over time. The central foci of the data are economic and demographic, with substantial detail on income sources and amounts, wealth, savings, employment, pensions, family composition changes, childbirth and marriage histories, and residential location. Over the life of the PSID, the NIA has funded supplements on wealth, health, parental health and long term care, housing, and the financial impact of illness, thus also making it possible to model retirement and residential mobility. Starting in 1999, much greater detail on specific health conditions and health care expenses is included for respondent and spouse. Other enhancements have included a question series about emotional distress (2001); the two stem questions from the Composite International Diagnostic Interview to assess symptoms of major depression (2003); a supplement on philanthropic giving and volunteering (2001-03); a question series on Internet and computer use (2003); linkage to the National Death Index with cause of death information for more than 4,000 individuals through the 1997 wave, updated for each subsequent wave; social and family history variables and GIS-linked environmental data; basic data on pension plans; event history calendar methodology to facilitate recall of employment spells (2001). The reporting unit is the family: single person living alone or sharing a household with other non-relatives; group of people related by blood, marriage, or adoption; unmarried couple living together in what appears to be a fairly permanent arrangement. Interviews were conducted annually from 1968 through 1997; biennial interviewing began in 1999. There is an oversample of Blacks (30%). Waves 1990 through 1995 included a 20% Hispanic oversample; within the Hispanic oversample, Cubans and Puerto Ricans were oversampled relative to Mexicans. All data from 1994 through 2001 are available as public release files; prior waves can be obtained in archive versions. The special files with weights for families are also available. Restricted files include the Geocode Match File with information for 1968 through 2001, the 1968-2001 Death File, and the 1991 Medicare Claims File. * Dates of Study: 1968-2003 * Study Features: Longitudinal, Minority Oversampling * Sample Size: 65,000+ Links * ICPSR Series: http://www.icpsr.umich.edu/icpsrweb/ICPSR/series/00131 * ICPSR 1968-1999: Annual Core Data: http://www.icpsr.umich.edu/icpsrweb/ICPSR/studies/07439 * ICPSR 1968-1999: Supplemental Files: http://www.icpsr.umich.edu/icpsrweb/ICPSR/studies/03202 * ICPSR 1989-1990: Latino Sample: http://www.icpsr.umich.edu/icpsrweb/ICPSR/studies/03203

Analysis of ‘GapMinder - Income Inequality’ provided by Analyst-2 (analyst-2.ai), based on source dataset retrieved from https://www.kaggle.com/psterk/income-inequality on 28 January 2022.

--- Dataset description provided by original source is as follows ---

Content

This analysis focuses on income inequailty as measured by the Gini Index* and its association with economic metrics such as GDP per capita, investments as a % of GDP, and tax revenue as a % of GDP. One polical metric, EIU democracy index, is also included.

The data is for years 2006 - 2016

This investigation can be considered a starting point for complex questions such as:

1. Is a higher tax revenue as a % of GDP associated with less income inequality?
2. Is a higher EIU democracy index associated with less income inequality?
3. Is higher GDP per capita associated with less income inequality?
4. Is higher investments as a % of GDP associated with less income inequality?

This analysis uses the gapminder dataset from the Gapminder Foundation. The Gapminder Foundation is a non-profit venture registered in Stockholm, Sweden, that promotes sustainable global development and achievement of the United Nations Millennium Development Goals by increased use and understanding of statistics and other information about social, economic and environmental development at local, national and global levels.

*The Gini Index is a measure of statistical dispersion intended to represent the income or wealth distribution of a nation's residents, and is the most commonly used measurement of inequality. It was developed by the Italian statistician and sociologist Corrado Gini and published in his 1912 paper Variability and Mutability.

The dataset contains data from the following GapMinder datasets:

EIU Democracy Index:

"This democracy index is using the data from the Economist Inteligence Unit to express the quality of democracies as a number between 0 and 100. It's based on 60 different aspects of societies that are relevant to democracy universal suffrage for all adults, voter participation, perception of human rights protection and freedom to form organizations and parties. The democracy index is calculated from the 60 indicators, divided into five ""sub indexes"", which are:

1. Electoral pluralism index;
2. Government index;
3. Political participation indexm;
4. Political culture index;
5. Civil liberty index.

The sub-indexes are based on the sum of scores on roughly 12 indicators per sub-index, converted into a score between 0 and 100. (The Economist publishes the index with a scale from 0 to 10, but Gapminder has converted it to 0 to 100 to make it easier to communicate as a percentage.)" https://docs.google.com/spreadsheets/d/1d0noZrwAWxNBTDSfDgG06_aLGWUz4R6fgDhRaUZbDzE/edit#gid=935776888

Income: GDP per capita, constant PPP dollars

GDP per capita measures the value of everything produced in a country during a year, divided by the number of people. The unit is in international dollars, fixed 2011 prices. The data is adjusted for inflation and differences in the cost of living between countries, so-called PPP dollars. The end of the time series, between 1990 and 2016, uses the latest GDP per capita data from the World Bank, from their World Development Indicators. To go back in time before the World Bank series starts in 1990, we have used several sources, such as Angus Maddison. https://www.gapminder.org/data/documentation/gd001/

Investments (% of GDP)

Capital formation is a term used to describe the net capital accumulation during an accounting period for a particular country. The term refers to additions of capital goods, such as equipment, tools, transportation assets, and electricity. Countries need capital goods to replace the older ones that are used to produce goods and services. If a country cannot replace capital goods as they reach the end of their useful lives, production declines. Generally, the higher the capital formation of an economy, the faster an economy can grow its aggregate income.

Tax revenue (% of GDP)

refers to compulsory transfers to the central governement for public purposes. Does not include social security. https://data.worldbank.org/indicator/GC.TAX.TOTL.GD.ZS

Context

Gapminder is an independent Swedish foundation with no political, religious or economic affiliations. Gapminder is a fact tank, not a think tank. Gapminder fights devastating misconceptions about global development. Gapminder produces free teaching resources making the world understandable based on reliable statistics. Gapminder promotes a fact-based worldview everyone can understand. Gapminder collaborates with universities, UN, public agencies and non-governmental organizations. All Gapminder activities are governed by the board. We do not award grants. Gapminder Foundation is registered at Stockholm County Administration Board. Our constitution can be found here.

Acknowledgements

Thanks to gapminder.org for organizing the above datasets.

Inspiration

Below are some research questions associated with the data and some initial conclusions:

Research Question 1 - Is Income Inequality Getting Worse or Better in the Last 10 Years?

Answer:

Yes, it is getting better, improving from 38.7 to 37.3

On a continent basis, all were either declining or mostly flat, except for Africa.

Research Question 2 - What Top 10 Countries Have the Lowest and Highest Income Inequality?

Answer:

Lowest: Slovenia, Ukraine, Czech Republic, Norway, Slovak Republic, Denmark, Kazakhstan, Finland, Belarus,Kyrgyz Republic

Highest: Colombia, Lesotho, Honduras, Bolivia, Central African Republic, Zambia, Suriname, Namibia, Botswana, South Africa

Research Question 3 Is a higher tax revenue as a % of GDP associated with less income inequality?

Answer: No

Research Question 4 - Is Higher Income Per Person - GDP Per Capita associated with less income inequality?

Answer: No, but weak negative correlation.

Research Question 5 - Is Higher Investment as % GDP associated with less income inequality?

Answer: No

Research Question 6 - Is Higher EIU Democracy Index associated with less income inequality?

Answer: No, but weak negative correlation.

The above results suggest that there are other drivers for the overall reduction in income inequality. Futher analysis of additional factors should be undertaken.

--- Original source retains full ownership of the source dataset ---

analyze the survey of income and program participation (sipp) with r if the census bureau's budget was gutted and only one complex sample survey survived, pray it's the survey of income and program participation (sipp). it's giant. it's rich with variables. it's monthly. it follows households over three, four, now five year panels. the congressional budget office uses it for their health insurance simulation . analysts read that sipp has person-month files, get scurred, and retreat to inferior options. the american community survey may be the mount everest of survey data, but sipp is most certainly the amazon. questions swing wild and free through the jungle canopy i mean core data dictionary. legend has it that there are still species of topical module variables that scientists like you have yet to analyze. ponce de león would've loved it here. ponce. what a name. what a guy. the sipp 2008 panel data started from a sample of 105,663 individuals in 42,030 households. once the sample gets drawn, the census bureau surveys one-fourth of the respondents every four months, over f our or five years (panel durations vary). you absolutely must read and understand pdf pages 3, 4, and 5 of this document before starting any analysis (start at the header 'waves and rotation groups'). if you don't comprehend what's going on, try their survey design tutorial. since sipp collects information from respondents regarding every month over the duration of the panel, you'll need to be hyper-aware of whether you want your results to be point-in-time, annualized, or specific to some other period. the analysis scripts below provide examples of each. at every four-month interview point, every respondent answers every core question for the previous four months. after that, wave-specific addenda (called topical modules) get asked, but generally only regarding a single prior month. to repeat: core wave files contain four records per person, topical modules contain one. if you stacked every core wave, you would have one record per person per month for the duration o f the panel. mmmassive. ~100,000 respondents x 12 months x ~4 years. have an analysis plan before you start writing code so you extract exactly what you need, nothing more. better yet, modify something of mine. cool? this new github repository contains eight, you read me, eight scripts: 1996 panel - download and create database.R 2001 panel - download and create database.R 2004 panel - download and create database.R 2008 panel - download and create database.R since some variables are character strings in one file and integers in anoth er, initiate an r function to harmonize variable class inconsistencies in the sas importation scripts properly handle the parentheses seen in a few of the sas importation scripts, because the SAScii package currently does not create an rsqlite database, initiate a variant of the read.SAScii function that imports ascii data directly into a sql database (.db) download each microdata file - weights, topical modules, everything - then read 'em into sql 2008 panel - full year analysis examples.R< br /> define which waves and specific variables to pull into ram, based on the year chosen loop through each of twelve months, constructing a single-year temporary table inside the database read that twelve-month file into working memory, then save it for faster loading later if you like read the main and replicate weights columns into working memory too, merge everything construct a few annualized and demographic columns using all twelve months' worth of information construct a replicate-weighted complex sample design with a fay's adjustment factor of one-half, again save it for faster loading later, only if you're so inclined reproduce census-publish ed statistics, not precisely (due to topcoding described here on pdf page 19) 2008 panel - point-in-time analysis examples.R define which wave(s) and specific variables to pull into ram, based on the calendar month chosen read that interview point (srefmon)- or calendar month (rhcalmn)-based file into working memory read the topical module and replicate weights files into working memory too, merge it like you mean it construct a few new, exciting variables using both core and topical module questions construct a replicate-weighted complex sample design with a fay's adjustment factor of one-half reproduce census-published statistics, not exactly cuz the authors of this brief used the generalized variance formula (gvf) to calculate the margin of error - see pdf page 4 for more detail - the friendly statisticians at census recommend using the replicate weights whenever possible. oh hayy, now it is. 2008 panel - median value of household assets.R define which wave(s) and spe cific variables to pull into ram, based on the topical module chosen read the topical module and replicate weights files into working memory too, merge once again construct a replicate-weighted complex sample design with a...

Abstract

The survey was conducted during December 2006, following an initial mini census listing exercise which was conducted about two months earlier in late September 2006. The objectives of the HIES were as follows: a) Provide information on income and expenditure distribution within the population; b) Provide income estimates of the household sector for the national accounts; c) Provide data for the re-base on the consumer price index; d) Provide data for the analysis of poverty and hardship.

Geographic coverage

National coverage: whole island was covered for the survey.

Analysis unit

• Household;
• Individual.

Universe

The survey covered all private households on the island of Nauru. When the survey was in the field, interviewers were further required to reduce the scope by removing those households which had not been residing in Nauru for the last 12 months and did not intend to stay in Nauru for the next 12 months. Persons living in special dwellings (Hospital, Prison, etc) were not included in the survey.

Kind of data

Sample survey data [ssd]

Sampling procedure

The sample size adopted for the survey was 500 households which allowed for expected sample loss, whilst still maintaining a suitable responding sample for the analysis.

Before the sample was selected, the population was stratified by constituency in order to assist with the logistical issues associated with the fieldwork. There were eight constituencies in total, along with "Location" which stretches across the districts of Denigamodu and Aiwo, forming nine strata in total. Although constituency level analysis was not a priority for the survey, sample sizes within each stratum were kept to a minimum of 40 households, to enable some basic forms of analysis at this level if required.

The sample selection procedure within each stratum was then to sort each household on the frame by household size (number of people), and then run a systematic skip through the list in order to achieve the desirable sample size.

Sampling deviation

No deviations from the sample design took place.

Mode of data collection

Face-to-face [f2f]

Research instrument

The survey schedules adopted for the Household Income and Expenditure Survey (HIES) included the following: · Expenditure questionnaire; · Income questionnaire; · Miscellaneous questionnaire; · Diary (x2).

Whilst a Household Control Form collecting basic demographics is also normally included with the survey, this wasn't required for this HIES as this activity took place for all households in the mini census.

Information collected in the four schedules covered the following: -Expenditure questionnaire: Covers basic details about the dwelling structure and its access to things like water and sanitation. It was also used as the vehicle to collect expenditure on major and infrequent expenditures incurred by the household. -Income questionnaire: Covers each of the main types of household income generated by the household such as wages and salaries, business income and income from subsistence activities. -Miscellaneous questionnaire: Covers topics relating to health access, labour force status and education. -Diary: Covers all day to day expenditures incurred by the household, consumption of items produced by the household such as fish and crops, and gifts both received and given by the household.

All questionnaires are provided as External Resources.

Cleaning operations

There were 3 phases to the editing process for the 2006 Household Income and Expenditure Survey (HIES) of Nauru which included: 1. Data Verification operations; 2. Data Editing operations; 3. Data Auditing operations.

The software used for data editting is CSPro 3.0. After each batch is completed the supervisor should check that all person details have been entered from the household listing form (HCF) and should review the income and expenditure questionnaires for each batch ensuring that all items have been entered correctly. Any omitted or incorrect items should be entered into the system. The supervisor is required to perform outlier checks (large or small values) on the batched diary data by calculating unit price (amount/quantity) and comparing prices for each item. This is to be conducted by loading the data into Excel files and sorting data by unit price for each item. Any changes to prices or quantities will be made on the batch file.

For more information on what each phase entailed go the document HIES Processing Instructions attached to this documentation.

Response rate

The survey response rates were a lot lower than expected, especially in some districts. The district of Aiwo, Uaboe and Denigomodu had the lowest response rates with 16.7%, 20.0% and 34.8% respectively. The area of Location was also extremely low with a responses rate of 32.2%. On a more positive note, the districts of Yaren, Ewa, Anabar, Ijuw and Anibare all had response rates at 80.0% or better.

The major contributing factor to the low response rates were households refusing to take part in the survey. The figures for responding above only include fully responding households, and given there were many partial responses, this also brought the values down. The other significant contributing factor to the low response rates was the interviewers not being able to make contact with the household during the survey period.

Unfortunately, not only do low response rates often increase the sampling error of the survey estimates, because the final sample is smaller, it will also introduce response bias into the final estimates. Response bias takes place when the households responding to the survey possess different characteristics to the households not responding, thus generating different results to what would have been achieved if all selected households responded. It is extremely difficult to measure the impact of the non-response bias, as little information is generally known about the non-responding households in the survey. For the Nauru 2006 HIES however, it was noted during the fieldwork that a higher proportion of the Chinese population residing in Nauru were more likely to not respond. Given it is expected their income and expenditure patterns would differ from the rest of the population, this would contribute to the magnitude of the bias.

Below is the list of all response rates by district: -Yaren: 80.5% -Boe: 70% -Aiwo: 16.7% -Buada: 62.5% -Denigomodu: 34.8% -Nibok: 68.4% -Uaboe: 20% -Baitsi: 47.8% -Ewa: 80% -Anetan: 76.5% -Anabar: 81.8% -Ijuw: 85.7% -Anibare: 80% -Meneng: 64.3% -Location: 32.2% -TOTAL: 54.4%

Sampling error estimates

To determine the impact of sampling error on the survey results, relative standard errors (RSEs) for key estimates were produced. When interpreting these results, one must remember that these figures don't include any of the non-sampling errors discussed in other sections of this documentation

To also provide a rough guide on how to interpret the RSEs provided in the main report, the following information can be used:

Category  Description
RSE < 5%  Estimate can be regarded as very reliable
5% < RSE < 10% Estimate can be regarded as good and usable
10% < RSE < 25% Estimate can be considered usable, with caution
RSE > 25%  Estimate should only be used with extreme caution

The actual RSEs for the key estimates can be found in Section 4.1 of the main report

As can be seen from these tables, the estimates for Total Income and Total Expenditure from the Household Income and Expenditure Survey (HIES) can be considered to be very good, from a sampling error perspective. The same can also be said for the Wage and Salary estimate in income and the Food estimate in expenditure, which make up a high proportion of each respective group.

Many of the other estimates should be used with caution, depending on the magnitude of their RSE. Some of these high RSEs are to be expected, due to the expected degree of variability for how households would report for these items. For example, with Business Income (RSE 56.8%), most households would report no business income as no household members undertook this activity, whereas other households would report large business incomes as it's their main source of income.

Data appraisal

Other than the non-response issues discussed in this documentation, other quality issues were identified which included: 1) Reporting errors Some of the different aspects contributing to the reporting errors generated from the survey, with some examples/explanations for each, include the following:

a) Misinterpretation of survey questions: A common mistake which takes place when conducting a survey is that the person responding to the questionnaire may interpret a question differently to the interviewer, who in turn may have interpreted the question differently to the people who designed the questionnaire. Some examples of this for a Household Income and Expenditure Survey (HIES) can include people providing answers in dollars and cents, instead of just dollars, or the reference/recall period for an “income” or “expenditure” is misunderstood. These errors can often see reported amounts out by a factor of 10 or even 100, which can have major impacts on final results.

b) Recall problems for the questionnaire information: The majority of questions in both of the income and expenditure questionnaires require the respondent to recall what took place over a 12 month period. As would be expected, people will often forget what took place up to 12 months ago so some

In 2022, San Francisco had the highest median household income of cities ranking within the top 25 in terms of population, with a median household income in of 136,692 U.S. dollars. In that year, San Jose in California was ranked second, and Seattle, Washington third.

Following a fall after the great recession, median household income in the United States has been increasing in recent years. As of 2022, median household income by state was highest in Maryland, Washington, D.C., Utah, and Massachusetts. It was lowest in Mississippi, West Virginia, and Arkansas. Families with an annual income of 25,000 and 49,999 U.S. dollars made up the largest income bracket in America, with about 25.26 million households.

Data on median household income can be compared to statistics on personal income in the U.S. released by the Bureau of Economic Analysis. Personal income rose to around 21.8 trillion U.S. dollars in 2022, the highest value recorded. Personal income is a measure of the total income received by persons from all sources, while median household income is “the amount with divides the income distribution into two equal groups,” according to the U.S. Census Bureau. Half of the population in question lives above median income and half lives below. Though total personal income has increased in recent years, this wealth is not distributed throughout the population. In practical terms, income of most households has decreased. One additional statistic illustrates this disparity: for the lowest quintile of workers, mean household income has remained more or less steady for the past decade at about 13 to 16 thousand constant U.S. dollars annually. Meanwhile, income for the top five percent of workers has actually risen from about 285,000 U.S. dollars in 1990 to about 499,900 U.S. dollars in 2020.

Although the American Community Survey (ACS) produces population, demographic and housing unit estimates, for 2020, the 2020 Census provides the official counts of the population and housing units for the nation, states, counties, cities, and towns. For 2016 to 2019, the Population Estimates Program provides estimates of the population for the nation, states, counties, cities, and towns and intercensal housing unit estimates for the nation, states, and counties..Supporting documentation on code lists, subject definitions, data accuracy, and statistical testing can be found on the American Community Survey website in the Technical Documentation section.Sample size and data quality measures (including coverage rates, allocation rates, and response rates) can be found on the American Community Survey website in the Methodology section..Source: U.S. Census Bureau, 2016-2020 American Community Survey 5-Year Estimates.Data are based on a sample and are subject to sampling variability. The degree of uncertainty for an estimate arising from sampling variability is represented through the use of a margin of error. The value shown here is the 90 percent margin of error. The margin of error can be interpreted roughly as providing a 90 percent probability that the interval defined by the estimate minus the margin of error and the estimate plus the margin of error (the lower and upper confidence bounds) contains the true value. In addition to sampling variability, the ACS estimates are subject to nonsampling error (for a discussion of nonsampling variability, see ACS Technical Documentation). The effect of nonsampling error is not represented in these tables..When information is missing or inconsistent, the Census Bureau logically assigns an acceptable value using the response to a related question or questions. If a logical assignment is not possible, data are filled using a statistical process called allocation, which uses a similar individual or household to provide a donor value. The "Allocated" section is the number of respondents who received an allocated value for a particular subject..Between 2018 and 2019 the American Community Survey retirement income question changed. These changes resulted in an increase in both the number of households reporting retirement income and higher aggregate retirement income at the national level. For more information see Changes to the Retirement Income Question ..The categories for relationship to householder were revised in 2019. For more information see Revisions to the Relationship to Household item..The 2016-2020 American Community Survey (ACS) data generally reflect the September 2018 Office of Management and Budget (OMB) delineations of metropolitan and micropolitan statistical areas. In certain instances, the names, codes, and boundaries of the principal cities shown in ACS tables may differ from the OMB delineation lists due to differences in the effective dates of the geographic entities..Estimates of urban and rural populations, housing units, and characteristics reflect boundaries of urban areas defined based on Census 2010 data. As a result, data for urban and rural areas from the ACS do not necessarily reflect the results of ongoing urbanization..Explanation of Symbols:- The estimate could not be computed because there were an insufficient number of sample observations. For a ratio of medians estimate, one or both of the median estimates falls in the lowest interval or highest interval of an open-ended distribution.N The estimate or margin of error cannot be displayed because there were an insufficient number of sample cases in the selected geographic area. (X) The estimate or margin of error is not applicable or not available.median- The median falls in the lowest interval of an open-ended distribution (for example "2,500-")median+ The median falls in the highest interval of an open-ended distribution (for example "250,000+").** The margin of error could not be computed because there were an insufficient number of sample observations.*** The margin of error could not be computed because the median falls in the lowest interval or highest interval of an open-ended distribution.***** A margin of error is not appropriate because the corresponding estimate is controlled to an independent population or housing estimate. Effectively, the corresponding estimate has no sampling error and the margin of error may be treated as zero.

This survey consists of 6 parts: - migration: data on migration, i.e. immigration and emigration of households or individual; this information is important for population forecasts and the evaluation of the development of the individual geographic area - handicaps: reasons for including questions on this topic are the “year of handicapped persons” (1981), proclaimed by the UN and paying attentions to these questions from a statistical point of view; the questions had already been posed in September 1987 (Mikrozensus MZ7803) - additional occupation: these questions should give information on additional occupation of employed and unemployed persons; of course the Mikrozensus can only document legal additional occupation, not illegal employment - social stratification: questions on occupational stratum and the receiving of benefit payments serve the in-depth analysis of the other questions - income: the currently available income data in Austria do not allow the representation of the population according to the total income of one person and according to the income of the household; a detailed income survey is not possible in the Mikrozensus: the question program on income is limited to a single question and self-employed, as well as persons helping in the family business are not interviewed. - birth-biography and desire to have children: these questions were for the most part already posed in June 1967 (Mikrozensus MZ7602).

Abstract

Tonga Household Income and Expenditure Survey 2009 (HIES), undertaken by the Tonga Statistics Department during the period from 1 January 2009 to 31 December 2009. This is the second survey of its kind in Tonga. The last one was carried out in 2000/01, and the results were used in November 2002 to rebase the Consumer Price Index (CPI). A report from that survey was produced in December 2002, and where possible, results from this report will be made to be comparable to the previous report.

• To provide updated information for the expenditure item weights for the CPI;

• To provide some data for the components of National Accounts; and

• To provide information on the nature and distribution of household income and expenditure for planners, policy makers, and the general public.

Geographic coverage

National Coverage and Island Division.

Analysis unit

Private Households, individuals, Income and expenditure items.

Universe

The survey covered all members of the household.

Kind of data

Sample survey data [ssd]

Sampling procedure

The sample design was done in such a way that promoted estimates primarily at the national level, but also at the island division level. For that reason a higher sample fraction was selected in the smaller island divisions.

Rural Tongatapu received the smallest sample fraction (8.3%) as it had the highest population. On the other hand the Ongo Niua received the largest sample fraction (21.5%) as their population was the smallest. Overall a sample of roughly 10 per cent was selected for Tonga.

The sample was selected independently within each of the 6 target areas. Firstly, extremely remote areas were removed from the frame (and thus not given a chance of selection) as it was considered too expensive to cover these areas. These areas only represented about 3.5 per cent of the total population for Tonga, so the impact of their removal was considered very minimal.

The sampling in each area was then undertaken using a two-stage process. The first stage involved the selection of census blocks using Probability Proportional to Size (PPS) sampling, where the size measure was the expected number of households in that block. For the second stage, a fixed number (twelve) of households were selected from each selected census block using systematic sampling. The household lists for all selected blocks were updated just prior to the second stage of selection.

Given the sample was spread out over four quarters during the 2009 calendar year, every 4th selected census block was allocated to a respective quarter. To ensure an equally distribution of sample to each quarter, the number of census blocks selected for each of the six target group was made divisible by four. This therefore meant the sample size for each target group was adjusted so that it was divisible by (4*12)=48, as can be seen in Table 1 of Section 1 of the survey report.

Mode of data collection

Face-to-face [f2f]

Research instrument

There were 4 main survey schedules used to collect the information for the survey were published in English: 1) Household Questionnaire 2) Individual Questionnaire - Part 1 3) Individual Questionnaire - Part 2 4) Individual Diary (x2)

Household Questionnaire

This questionnaire is primarily used to collect information on large expenditure items, but also collects information about the dwelling characteristics. In total there are 14 sections to this uestionnaire which cover: 1 Dwelling Characteristics 2 Household Possessions 3 Dwelling Tenure 4 Construction of Dwellings 5 Household Bills 6 Transport Expenses 7 Major Consumer Durables 8 Education/Recreation 9 Medical & Health 10 Overseas Travel 11 Special Events 12 Subsistence Activity Sales 13 Remittances 14 Contributions to Church/Village/School As stated above, the first section is devoted to collecting information about key dwelling characteristics, whereas the second section collects information on household possessions. Sections 3-11, and Section 14, focus on expenses the household incurs, whereas Section 13 focuses on remittances both paid by and received by the household. Finally, Section 12 collects information from households about the income they generate from subsistence activities. This section is the main question collecting income from the household questionnaire, as was included here as it was considered more appropriate to collect this data at the household level. The front page of this Questionnaire is also used for collecting the Roster of Household Members.

Individual Questionnaire - Part 1

This questionnaire collects basic demographic information about each individual in the household, including: • Relationship to Household Head • Sex • Age • Ethnicity • Marital Status

Also collected in this form is information about health problems each individual may have encountered in the last 3 months, followed by education information. For the education section, if a person is currently attending an education institution, then current level is asked, whereas if the person attended an education institution but no longer attends, then the highest level completed is collected. The last main section of this form collects information about labour force and is only asked of individuals aged 10 years and above. These questions aim to classify each person in scope for this section as either: • In the Labour Force - Employed • In the Labour Force - Unemployed • Not in the Labour Force

Individual Questionnaire - Part 2

This questionnaire is focused on collecting information from individuals regarding their income. There are eight sections to this questionnaire of which six are devoted to income. They include: 1 Wages and Salary 2 Self-Employment
3 Previous Jobs
4 Ad-hoc Jobs 5 Pensions/Welfare Benefits 6 Other Income 7 Loan Information 8 Contributions to Benefit Schemes

As stated above, the first six sections of this questionnaire focus on income. Section 7 collects information pertaining to loans for i) households, ii) cars, iii) special events and iv) other, and finally the last question is an expense related question covering contributions to benefit schemes which was considered best covered at an individual level.

Individual Diary

The last form used for the survey was the Individual Diary which each individual aged 10 years and over was required to fill in for two weeks (two one-week diaries).

Each diary had 4 sections covering the following: 1) Items Purchased: This section had a separate page for each day and was for recording all items bought in a store, street vendors, market or any other place (including credit) 2) Home Grown/Produced Items: This section was for recording home grown/produced items consisting of items such as food grown at home or at the family plantation, self caught or gathered fish and homemade handicrafts and other goods grown and produced at home. Information is recorded for these items consumed by the household which they produced themselves, these items they gave away as a gift, and these items they received as a gift. 3) Gifts Given and Received: This section of the diary is for recording gifts given and received including both cash and purchased goods (but not home produced). If any member of the household receives a gift that meets this criteria during the diary keeping period from someone who is not a member of their household it is recorded here. 4) Winnings from Gambling: The last section of the Diary is for recording all winnings from gambling during the diary keeping period.

Cleaning operations

Batch edits in CSPro were performed on the data after data entry was completed. The batch edits were aimed at identifying any values falling outside acceptable ranges, as well as other inconsistencies in the data. As this process was done at the batch level, questionnaires were often referred to and manual changes to the data were performed to amend identified errors.

One significant problem which was identified during this process was the incorrect coding of phone card purchase to the purchase of actual phones. As there were many such cases, an automatic code change was applied to any purchase of phones which was less than $40 - recoding them to purchase of phone cards.

Response rate

The final Response Rates for the survey was high, which will assist in yielding statistically significant estimates. Across all six target groups the response rate was in excess of 95 per cent, with the exception of Ongo Niua who only reported 50 per cent. The reason the number was so low in the Ongo Niua was because this target area was only visited in the 2nd quarter, where half the total sample were enumerated (to make up for the sample loss in the first quarter), and was not visited again in quarter 3 and 4.

The reason behind the high response rates in other areas was due to the updated lists for selected census blocks excluding vacant dwellings. As such, it was mostly refusals that impacted on the final response rates.

Sampling error estimates

Sampling errors refer to those errors that are implicit in any sample survey, where only a portion of the population is covered. Non-sampling errors refer to all other types of error. These can arise at any stage of the survey process. Examples of activities that are likely to increase the level of non-sampling error are: failing to select a proper sample, poor questionnaire design, weak field supervision, inaccurate data entry, insufficient data editing, or failure to analyze or report on the data

Abstract

The main purpose of a HIES survey was to present high quality and representative national household data on income and expenditure in order to update Consumer Price Index (CPI), improve statistics on National Accounts and measure poverty within the country. These statistics are a requirement for evidence based policy-making in reducing poverty within the country and monitor progress in the national strategic plan "Te Kakeega 3".

The 2015-16 Household Income and Expenditure Survey (HIES) is the third HIES that was conducted by the Central Statistics Division since Tuvalu gained political independence in 1978. With great assitance from the Pacific Community (SPC) experts, the HIES was conducted over a period of 12 months in urban (Funafuti) and rural (4 outer islands) areas. From a total of 1,872 households on Tuvalu, an amount of 38 percent sample of all households in Tuvalu was selected to provide valid response.

Geographic coverage

National Coverage.

Analysis unit

Household and Individual.

Universe

The scope of the 2015/2016 Household Income and Expenditure Survey (HIES) was all occupied households in Tuvalu. Households are the sampling unit, defined as a group of people (related or not) who pool their money, and cook and eat together. It is not the physical structure (dwelling) in which people live. HIES covered all persons who were considered to be usual residents of private dwellings (must have been living in Tuvalu for a period of 12-months, or have intention to live in Tuvalu for a period of 12-months in order to be included in the survey). Usual residents who are temporary away are included as well (e.g., for work or a holiday).

Kind of data

Sample survey data [ssd]

Sampling procedure

Out of the total 1,872 households (HHs) listed in 2015, a sample 706 households which is 38 percent of the the total households were succesfully interviewed for a response rate of 98%.

SAMPLING FRAME: The 2010 (Household Income and Expenditure Survey (HIES) sample was spread over 12 months rounds - one each quarter - and the specifications of the final responding households are summarised below: Tuvalu urban: Selected households: 259 = 217 responded; Tuvalu rural: Selected households: 346 = 324 responded.

In 2010, 605 HHs were selected and 541 sufficiently responded. The 2010 HIES provided solid estimates for expenditure aggregates at the national level (sampling error for national expenditure estimate is 3.1%).

Similarly to the 2010 HIES, private occupied dwellings were the statistical unit for the 2015/2016 HIES. Institutions and vacant dwellings were removed from the sampling frame. Some areas in Tuvalu are very difficult to reach due to the cost of transportation and the remoteness of some islands, which is why they are excluded from the sample selection. The following table presents the distribution of the HHs according to their location (main island or outer islands in each domain) based on the 2012 Population and Housing Census: -Urban - Funafuti: 845 (48%); -Rural - Nanumea: 115 (7%); -Rural - Nanumaga: 116 (7%); -Rural - Niutao: 123 (7%); -Rural - Nui: 138 (8%); -Rural - Vaitupu: 226 (13%); -Rural - Nukufetau: 124 (%); -Rural - Nukulaelae: 67 (%); -Rural - Niulakita: 7 (%); -TOTAL: 1761 (100%).

The 2012 Population and Household Census (PHC) wsa used to select the island to interview, and then in each selected island the HH listing was updated for selection. For budget and logistics reasons the islands of Nui, Nukufetau, Nukulaelae and Niukalita were excluded from the sample selection. In total 19% of the HHs were excluded from the selection but this decision should not affect the HIES outputs as those 19% show similar profile as other HHs who live in the outer islands. This exclusion will be take into consideration in the sampling weight computation in order to cover 100% of the outer island HHs.

SAMPLE SELECTION AND SAMPLE SIZE: A simple random selection was used in each of the selected island (HHs were selected directly from the sampling frame). Based on the findings from the 2010 Tuvalu HIES, the sample in Funafuti has been increased and the one in rural remains stable. Within each rural selected atolls, the allocation of the sample size is proportional to its size (baed on the 2012 population census). The table below shows the number of HHs to survey: Urban - Funafuti: 384; Rural - Vaitupu: 126; Rural - Nanumea: 63; Rural - Niutao: 84; Rural - Nanumaga: 63; TUVALU: 720.

The expected sample size has been increased by one third (361 HHs) with the aim of pre-empting the non contacted HHs (refusals, absence….). The 2015/2016 HIES adopted the standardized HIES methodology and survey instruments for the Pacific Islands region. This approach, developed by the Pacific Community (SPC), has resulted in proven survey forms being used for data collection. It involves collection of data over a 12-month period to account for seasonal changes in income and expenditure patterns, and to keep the field team to a smaller and more qualified group. Their implementation had the objective of producing consistent and high quality data.

Sampling deviation

For budget and logistics reasons the islands of Nui, Nukufetau, Nukulaelae and Niukalita were excluded from the sample selection. In total 19% of the HHs were excluded from the selection but this decision should not affect the HIES outputs as those 19% show similar profile as other HHs who live in the outer islands. This exclusion will be take into consideration in the sampling weight computation in order to cover 100% of the outer island HHs.

Mode of data collection

Face-to-face [f2f]

Research instrument

The survey contain 4 modules and 2 Diaries (1 diary for each of the two weeks that a household was enumerated). The purpose of a Diary is to record all the daily expenses and incomes of a Household as shown by its topics below; - DIARY
The Diary module contains questions such as "What did your Household buy Today (Food and Non-Food Items)?", "Payments for Services made Today", "Food, Non-Food and Services Received for Free", "Home-Produced Items Today", "Overflow Sheet for Items Bought This Week", "Overflow Sheet for Services Paid for This Week", "Overflow Sheet for Items Received for Free this Week", and an "Overflow Sheet for Home-Produced Items This Week".

The 4 modules are detailed below; - MODULE 1 - DEMOGRAPHIC INFORMATION The module contains individual demograhic questions on their Demographic Profiles, Labour Force status (Activities), Education status, Health status, Communication status and questions on "Household members that have left the household". - MODULE 2 - HOUSEHOLD EXPENDITURE The module contains household expenditure questions the housing characteristics, Housing tenure expenditures, Utilities and Communication, Land, Household goods and assets, Vehicles and accessories, Private Travel details, Household services expenditures, Cash contributions, Provisions of Financial support, Household asset insurance and taxes and questions on Personal insurance. - MODULE 3 - INDIVIDUAL EXPENDITURE This module contains individual expenditure questions on Education, Health, Clothing, Communication, Luxury Items, Alcohol, Kava and Tobacco, and Deprivation questions. - MODULE 4 - HOUSEHOLD & INDIVIDUAL INCOME
This module contains household and individual questions on their income, on topics such as Wages and Salary, Agricultural and Forestry Activities, Fishing, Gathering and Hunting Activities, Livestock and Aquaculture Activities, Handicraft/Home-processed Food Activities, Income from Non-subsistence Business, Property income, transfer income & other Receipts, and Remmitances and other Cash gifts.

Depending on the information being collected, a recall period (ranging from the last 7 days to the last 12 months) is applied to various sections of the questionnaire. The forms were completed by face-to-face interview, usually with the HH head providing most of the information, with other household (HH) members being interviewed when necessary. The interviews took place over a 2-week period such that the HH diary, which is completed by the HH on a daily basis for 2 weeks, can be monitored while the module interviews take place. The HH diary collects information on the HH's daily expenditure on goods and services; and the harvest, capture, collection or slaughter of primary produce (fruit, vegetables and animals) by intended purpose (home consumption, sale or to give away). The income and expenditure data from the modules and the diary are concatenated (ensuring that double counting does not occur), annualised, and extrapolated to form the income and expenditure aggregates presented herein.

Cleaning operations

The survey procedure and enumeration team structure allowed for in-round data entry, which gives the field staff the opportunity to correct the data by manual review and by using the entry system-generated error messages. This process was designed to improve data quality. The data entry system used system-controlled entry, interactive coding and validity and consistency checks. Despite the validity and consistency checks put in place, the data still required cleaning. The cleaning was a 2-stage process, which included manual cleaning while referencing the questionnaire, whereas the second stage involved computer-assisted code verification and, in some cases, imputation. Once the data were clean, verified and consistent, they were recoded to form a final aggregated database, consisting of: 1. Person level record - characteristics of every HH member, including activity

Abstract

Tonga Household Income and Expenditure Survey 2009 (HIES), undertaken by the Tonga Statistics Department during the period from 1 January 2009 to 31 December 2009. This is the second survey of its kind in Tonga. The last one was carried out in 2000/01, and the results were used in November 2002 to rebase the Consumer Price Index (CPI). A report from that survey was produced in December 2002, and where possible, results from this report will be made to be comparable to the previous report.

• To provide updated information for the expenditure item weights for the CPI; • To provide some data for the components of National Accounts; and • To provide information on the nature and distribution of household income and expenditure for planners, policy makers, and the general public.

Geographic coverage

National Coverage and Island Division

Analysis unit

Private Household, individual, income and expenditure item

Kind of data

Sample survey data [ssd]

Sampling procedure

The sample design was done in such a way that promoted estimates primarily at the national level, but also at the island division level. For that reason a higher sample fraction was selected in the smaller island divisions.

Rural Tongatapu received the smallest sample fraction (8.3%) as it had the highest population. On the other hand the Ongo Niua received the largest sample fraction (21.5%) as their population was the smallest. Overall a sample of roughly 10 per cent was selected for Tonga.

The sample was selected independently within each of the 6 target areas. Firstly, extremely remote areas were removed from the frame (and thus not given a chance of selection) as it was considered too expensive to cover these areas. These areas only represented about 3.5 per cent of the total population for Tonga, so the impact of their removal was considered very minimal.

The sampling in each area was then undertaken using a two-stage process. The first stage involved the selection of census blocks using Probability Proportional to Size (PPS) sampling, where the size measure was the expected number of households in that block. For the second stage, a fixed number (twelve) of households were selected from each selected census block using systematic sampling. The household lists for all selected blocks were updated just prior to the second stage of selection.

Given the sample was spread out over four quarters during the 2009 calendar year, every 4th selected census block was allocated to a respective quarter. To ensure an equally distribution of sample to each quarter, the number of census blocks selected for each of the six target group was made divisible by four. This therefore meant the sample size for each target group was adjusted so that it was divisible by (4*12)=48, as can be seen in Table 1 of Section 1 of the survey report.

Mode of data collection

Face-to-face [f2f]

Research instrument

There were 4 main survey schedules used to collect the information for the survey: 1) Household Questionnaire 2) Individual Questionnaire - Part 1 3) Individual Questionnaire - Part 2 4) Individual Diary (x2)

Household Questionnaire This questionnaire is primarily used to collect information on large expenditure items, but also collects information about the dwelling characteristics. In total there are 14 sections to this questionnaire which cover: 1 Dwelling Characteristics 2 Household Possessions 3 Dwelling Tenure 4 Construction of Dwellings 5 Household Bills 6 Transport Expenses 7 Major Consumer Durables 8 Education/Recreation 9 Medical & Health 10 Overseas Travel 11 Special Events 12 Subsistence Activity Sales 13 Remittances 14 Contributions to Church/Village/School As stated above, the first section is devoted to collecting information about key dwelling characteristics, whereas the second section collects information on household possessions. Sections 3-11, and Section 14, focus on expenses the household incurs, whereas Section 13 focuses on remittances both paid by and received by the household. Finally, Section 12 collects information from households about the income they generate from subsistence activities. This section is the main question collecting income from the household questionnaire, as was included here as it was considered more appropriate to collect this data at the household level. The front page of this Questionnaire is also used for collecting the Roster of Household Members.

Individual Questionnaire - Part 1 This questionnaire collects basic demographic information about each individual in the household, including: • Relationship to Household Head • Sex • Age • Ethnicity • Marital Status

Also collected in this form is information about health problems each individual may have encountered in the last 3 months, followed by education information. For the education section, if a person is currently attending an education institution, then current level is asked, whereas if the person attended an education institution but no longer attends, then the highest level completed is collected. The last main section of this form collects information about labour force and is only asked of individuals aged 10 years and above. These questions aim to classify each person in scope for this section as either: • In the Labour Force - Employed • In the Labour Force - Unemployed • Not in the Labour Force

Individual Questionnaire - Part 2 This questionnaire is focused on collecting information from individuals regarding their income. There are eight sections to this questionnaire of which six are devoted to income. They include: 1 Wages and Salary 2 Self-Employment
3 Previous Jobs
4 Ad-hoc Jobs 5 Pensions/Welfare Benefits 6 Other Income 7 Loan Information 8 Contributions to Benefit Schemes

As stated above, the first six sections of this questionnaire focus on income. Section 7 collects information pertaining to loans for i) households, ii) cars, iii) special events and iv) other, and finally the last question is an expense related question covering contributions to benefit schemes which was considered best covered at an individual level.

Individual Diary The last form used for the survey was the Individual Diary which each individual aged 10 years and over was required to fill in for two weeks (two one-week diaries).

Each diary had 4 sections covering the following: 1) Items Purchased: This section had a separate page for each day and was for recording all items bought in a store, street vendors, market or any other place (including credit) 2) Home Grown/Produced Items: This section was for recording home grown/produced items consisting of items such as food grown at home or at the family plantation, self caught or gathered fish and homemade handicrafts and other goods grown and produced at home. Information is recorded for these items consumed by the household which they produced themselves, these items they gave away as a gift, and these items they received as a gift. 3) Gifts Given and Received: This section of the diary is for recording gifts given and received including both cash and purchased goods (but not home produced). If any member of the household receives a gift that meets this criteria during the diary keeping period from someone who is not a member of their household it is recorded here. 4) Winnings from Gambling: The last section of the Diary is for recording all winnings from gambling during the diary keeping period.

Cleaning operations

Batch edits in CSPro were performed on the data after data entry was completed. The batch edits were aimed at identifying any values falling outside acceptable ranges, as well as other inconsistencies in the data. As this process was done at the batch level, questionnaires were often referred to and manual changes to the data were performed to amend identified errors.

One significant problem which was identified during this process was the incorrect coding of phone card purchase to the purchase of actual phones. As there were many such cases, an automatic code change was applied to any purchase of phones which was less than $40 - recoding them to purchase of phone cards.

Response rate

The final Response Rates for the survey was high, which will assist in yielding statistically significant estimates. Across all six target groups the response rate was in excess of 95 per cent, with the exception of Ongo Niua who only reported 50 per cent. The reason the number was so low in the Ongo Niua was because this target area was only visited in the 2nd quarter, where half the total sample were enumerated (to make up for the sample loss in the first quarter), and was not visited again in quarter 3 and 4.

The reason behind the high response rates in other areas was due to the updated lists for selected census blocks excluding vacant dwellings. As such, it was mostly refusals that impacted on the final response rates.

Sampling error estimates

Sampling errors refer to those errors that are implicit in any sample survey, where only a portion of the population is covered. Non-sampling errors refer to all other types of error. These can arise at any stage of the survey process. Examples of activities that are likely to increase the level of non-sampling error are: failing to select a proper sample, poor questionnaire design, weak field supervision, inaccurate data entry, insufficient data editing, or failure to analyze or report on the data correctly. If a census of all the households in Tonga were carried out, there would be no sampling error (but probably increased non-sampling

Although the American Community Survey (ACS) produces population, demographic and housing unit estimates, it is the Census Bureau's Population Estimates Program that produces and disseminates the official estimates of the population for the nation, states, counties, cities, and towns and estimates of housing units for states and counties..Supporting documentation on code lists, subject definitions, data accuracy, and statistical testing can be found on the American Community Survey website in the Technical Documentation section.Sample size and data quality measures (including coverage rates, allocation rates, and response rates) can be found on the American Community Survey website in the Methodology section..Source: U.S. Census Bureau, 2017-2021 American Community Survey 5-Year Estimates.Data are based on a sample and are subject to sampling variability. The degree of uncertainty for an estimate arising from sampling variability is represented through the use of a margin of error. The value shown here is the 90 percent margin of error. The margin of error can be interpreted roughly as providing a 90 percent probability that the interval defined by the estimate minus the margin of error and the estimate plus the margin of error (the lower and upper confidence bounds) contains the true value. In addition to sampling variability, the ACS estimates are subject to nonsampling error (for a discussion of nonsampling variability, see ACS Technical Documentation). The effect of nonsampling error is not represented in these tables..Between 2018 and 2019 the American Community Survey retirement income question changed. These changes resulted in an increase in both the number of households reporting retirement income and higher aggregate retirement income at the national level. For more information see Changes to the Retirement Income Question ..Beginning in data year 2019, respondents to the Weeks Worked question provided an integer value for the number of weeks worked. For data years 2008 through 2018, respondents selected a category corresponding to the number of weeks worked..The 2017-2021 American Community Survey (ACS) data generally reflect the March 2020 Office of Management and Budget (OMB) delineations of metropolitan and micropolitan statistical areas. In certain instances, the names, codes, and boundaries of the principal cities shown in ACS tables may differ from the OMB delineation lists due to differences in the effective dates of the geographic entities..Estimates of urban and rural populations, housing units, and characteristics reflect boundaries of urban areas defined based on Census 2010 data. As a result, data for urban and rural areas from the ACS do not necessarily reflect the results of ongoing urbanization..Explanation of Symbols:- The estimate could not be computed because there were an insufficient number of sample observations. For a ratio of medians estimate, one or both of the median estimates falls in the lowest interval or highest interval of an open-ended distribution. For a 5-year median estimate, the margin of error associated with a median was larger than the median itself.N The estimate or margin of error cannot be displayed because there were an insufficient number of sample cases in the selected geographic area. (X) The estimate or margin of error is not applicable or not available.median- The median falls in the lowest interval of an open-ended distribution (for example "2,500-")median+ The median falls in the highest interval of an open-ended distribution (for example "250,000+").** The margin of error could not be computed because there were an insufficient number of sample observations.*** The margin of error could not be computed because the median falls in the lowest interval or highest interval of an open-ended distribution.***** A margin of error is not appropriate because the corresponding estimate is controlled to an independent population or housing estimate. Effectively, the corresponding estimate has no sampling error and the margin of error may be treated as zero.

Although the American Community Survey (ACS) produces population, demographic and housing unit estimates, it is the Census Bureau's Population Estimates Program that produces and disseminates the official estimates of the population for the nation, states, counties, cities, and towns and estimates of housing units for states and counties..Supporting documentation on code lists, subject definitions, data accuracy, and statistical testing can be found on the American Community Survey website in the Technical Documentation section.Sample size and data quality measures (including coverage rates, allocation rates, and response rates) can be found on the American Community Survey website in the Methodology section..Source: U.S. Census Bureau, 2019 American Community Survey 1-Year Estimates.Data are based on a sample and are subject to sampling variability. The degree of uncertainty for an estimate arising from sampling variability is represented through the use of a margin of error. The value shown here is the 90 percent margin of error. The margin of error can be interpreted roughly as providing a 90 percent probability that the interval defined by the estimate minus the margin of error and the estimate plus the margin of error (the lower and upper confidence bounds) contains the true value. In addition to sampling variability, the ACS estimates are subject to nonsampling error (for a discussion of nonsampling variability, see ACS Technical Documentation). The effect of nonsampling error is not represented in these tables..Between 2018 and 2019 the American Community Survey retirement income question changed. These changes resulted in an increase in both the number of households reporting retirement income and higher aggregate retirement income at the national level. For more information see Changes to the Retirement Income Question ..The 2019 American Community Survey (ACS) data generally reflect the September 2018 Office of Management and Budget (OMB) delineations of metropolitan and micropolitan statistical areas. In certain instances the names, codes, and boundaries of the principal cities shown in ACS tables may differ from the OMB delineations due to differences in the effective dates of the geographic entities..Estimates of urban and rural populations, housing units, and characteristics reflect boundaries of urban areas defined based on Census 2010 data. As a result, data for urban and rural areas from the ACS do not necessarily reflect the results of ongoing urbanization..Explanation of Symbols:An "**" entry in the margin of error column indicates that either no sample observations or too few sample observations were available to compute a standard error and thus the margin of error. A statistical test is not appropriate.An "-" entry in the estimate column indicates that either no sample observations or too few sample observations were available to compute an estimate, or a ratio of medians cannot be calculated because one or both of the median estimates falls in the lowest interval or upper interval of an open-ended distribution, or the margin of error associated with a median was larger than the median itself.An "-" following a median estimate means the median falls in the lowest interval of an open-ended distribution.An "+" following a median estimate means the median falls in the upper interval of an open-ended distribution.An "***" entry in the margin of error column indicates that the median falls in the lowest interval or upper interval of an open-ended distribution. A statistical test is not appropriate.An "*****" entry in the margin of error column indicates that the estimate is controlled. A statistical test for sampling variability is not appropriate. An "N" entry in the estimate and margin of error columns indicates that data for this geographic area cannot be displayed because the number of sample cases is too small.An "(X)" means that the estimate is not applicable or not available.

This data collection is part of a longitudinal survey designed to provide detailed information on the economic situation of households and persons in the United States. These data examine the distribution of income, wealth, and poverty in American society and gauge the effects of federal and state programs on the well-being of families and individuals. There are three basic elements contained in the survey. The first is a control card that records basic social and demographic characteristics for each person in a household, as well as changes in such characteristics over the course of the interviewing period. These include age, sex, race, ethnic origin, marital status, household relationship, education, and veteran status. Limited data are provided on housing unit characteristics such as units in structure, tenure, access, and complete kitchen facilities. The second element is the core portion of the questionnaire, with questions repeated at each interview on labor force activity, types and amounts of income, and participation in various cash and noncash benefit programs for each month of the four- month reference period. Data for employed persons include number of hours and weeks worked, earnings, and weeks without a job. Nonworkers are classified as unemployed or not in the labor force. In addition to providing income data associated with labor force activity, the core questions cover nearly 50 other types of income. Core data also include postsecondary school attendance, public or private subsidized rental housing, low-income energy assistance, and school breakfast and lunch participation. The third element consists of topical modules, which are a series of supplemental questions asked during selected household visits. Topical modules include some core data to link individuals to the core files. The Wave 1 Topical Module covers recipiency and employment history. The Wave 2 Topical Module includes work disability, education and training, marital, migration, and fertility histories, and household relationships. The Wave 3 Topical Module covers medical expenses and utilization of health care, work-related expenses and child support, assets and liabilities, real estate, shelter costs, dependent care, vehicles, value of business, interest earning accounts, rental properties, stocks and mutual fund shares, mortgages, and other assets. The Wave 4 Topical Module covers work schedule, taxes, child care, and annual income and retirement accounts. Data in the Wave 5 Topical Module describe child support agreements, school enrollment and financing, support for non-household members, adult and child disability, and employer-provided health benefits. The Wave 6 Topical Module covers medical expenses and utilization of health care, work related expenses, child support paid and child care poverty, assets and liabilities, real estate, shelter costs, dependent care, vehicles, value of business, interest earning accounts, rental properties, stock and mutual fund shares, mortgages, and other financial investments. The Wave 7 Topical Module covers informal caregiving, children's well-being, and annual income and retirement accounts. The Wave 8 Topical Module and Wave 8 Welfare Reform Topical Module cover child support agreements, support for nonhousehold members, adult disability, child disability, adult well-being, and welfare reform. The Wave 9 Topical Module covers medical expenses and utilization of heath care (adults and children), work related expenses, child support paid and child care poverty, assets and liabilities, real estate, shelter costs, dependent care, vehicles, value of business, interest earnings accounts, rental properties, stocks and mutual fund shares mortgages, and other financial investments (Source: downloaded from ICPSR 7/13/10)

Although the American Community Survey (ACS) produces population, demographic and housing unit estimates, it is the Census Bureau's Population Estimates Program that produces and disseminates the official estimates of the population for the nation, states, counties, cities, and towns and estimates of housing units for states and counties..Supporting documentation on code lists, subject definitions, data accuracy, and statistical testing can be found on the American Community Survey website in the Technical Documentation section.Sample size and data quality measures (including coverage rates, allocation rates, and response rates) can be found on the American Community Survey website in the Methodology section..Source: U.S. Census Bureau, 2017-2021 American Community Survey 5-Year Estimates.Data are based on a sample and are subject to sampling variability. The degree of uncertainty for an estimate arising from sampling variability is represented through the use of a margin of error. The value shown here is the 90 percent margin of error. The margin of error can be interpreted roughly as providing a 90 percent probability that the interval defined by the estimate minus the margin of error and the estimate plus the margin of error (the lower and upper confidence bounds) contains the true value. In addition to sampling variability, the ACS estimates are subject to nonsampling error (for a discussion of nonsampling variability, see ACS Technical Documentation). The effect of nonsampling error is not represented in these tables..Between 2018 and 2019 the American Community Survey retirement income question changed. These changes resulted in an increase in both the number of households reporting retirement income and higher aggregate retirement income at the national level. For more information see Changes to the Retirement Income Question ..The 2017-2021 American Community Survey (ACS) data generally reflect the March 2020 Office of Management and Budget (OMB) delineations of metropolitan and micropolitan statistical areas. In certain instances, the names, codes, and boundaries of the principal cities shown in ACS tables may differ from the OMB delineation lists due to differences in the effective dates of the geographic entities..Estimates of urban and rural populations, housing units, and characteristics reflect boundaries of urban areas defined based on Census 2010 data. As a result, data for urban and rural areas from the ACS do not necessarily reflect the results of ongoing urbanization..Explanation of Symbols:- The estimate could not be computed because there were an insufficient number of sample observations. For a ratio of medians estimate, one or both of the median estimates falls in the lowest interval or highest interval of an open-ended distribution. For a 5-year median estimate, the margin of error associated with a median was larger than the median itself.N The estimate or margin of error cannot be displayed because there were an insufficient number of sample cases in the selected geographic area. (X) The estimate or margin of error is not applicable or not available.median- The median falls in the lowest interval of an open-ended distribution (for example "2,500-")median+ The median falls in the highest interval of an open-ended distribution (for example "250,000+").** The margin of error could not be computed because there were an insufficient number of sample observations.*** The margin of error could not be computed because the median falls in the lowest interval or highest interval of an open-ended distribution.***** A margin of error is not appropriate because the corresponding estimate is controlled to an independent population or housing estimate. Effectively, the corresponding estimate has no sampling error and the margin of error may be treated as zero.

The Household Market and Nonmarket Activities (HUS) project started as a joint research project between the Industrial Institute for Economic and Social Research (IUI) and Göteborg University in 1980. The ambition was to build a consistent longitudinal micro data base on the use of time, money and public services of households. The first main survey was carried out in 1984. In addition to a contact interview with the selected individuals, all designated individuals participated in a personal interview and two telephone interviews. All respondents were asked about their family background, education, marital status, labor market experience, and employment. In addition, questions about the household were asked of the head of household, concerning family composition, child care, health status, housing, possession of vacation homes, cars, boats and other consumption durables. At the end of the personal interview the household head had to fill out a questionnaire including questions about financing of current home, construction costs for building a house, house value and loans, imputation of property values and loans, additions/renovations 1983, maintenance and repairs, leasing, sale of previous home, assets and liabilities, and non-taxable benefits. All the respondents had to fill out a questionnaire including questions about tax-return information 1983, employment income, and taxes and support payments. Two telephone interviews were used primarily to collect data on the household´s time use and consumption expenditures. The 1986 HUS-survey included both a follow-up of the 1984 sample (panel study) and a supplementary sample. The 1986 sample included 1) all respondents participating in the 1984 survey, 2) the household heads, partners and third persons who should have participated in 1984 but did not (1984 nonresponse), 3) those individuals who started living together after the 1984 interview with an selected individual who participated or was supposed to participate in 1984, 4) members of the 1984 household born in 1966 or 1967. If entering a new household, for example because of leaving their parental home, the household head and his/her partner were also interviewed. Respondents participating in the 1984 survey were interviewed by telephone in 1986. Questions dealt with changes in family composition, housing, employment, wages and child care, and it was not only recorded whether a change had occurred, and what sort of change, but also when it occurred. The respondents also received a questionnaire by mail with questions mainly concerning income and assets. Respondents not participating in the earlier survey were interviewed in person and were asked approximately the same questions as in the 1984 personal interview. The 1988 HUS-survey was considerably smaller than the previous ones. It was addressed exclusively to participants in the 1986 survey, and consisted of a self-enumerated questionnaire with a nonrespondent follow-up by telephone. The questions dealt with changes in housing conditions, employment and household composition. The questionnaire also contained some questions on household income. In many respect the 1991 HUS-survey replicated the 1988 survey. The questions were basically the same in content and range, and the survey was conducted as a self-enamurated questionnaire sent out by mail. This time, however, in contrast to the 1988 survey, an attempt was made to include in the survey the new household members who had moved into sample households since 1986, as well as young people who turned 18 after the 1986 survey. Earlier respondents received a questionnaire by mail containing questions about their home, their primary occupation and weekly work hours since May 1988 (event-history data), earnings in 1989, 1990 and 1991, household composition and any changes in it that might have occurred since 1988, child care and some questions on income. New respondents were also asked about their education and labor-market experience. With respect to its design and question wording, the 1993 survey is a new version of the 1986 survey. The survey is made up of four parts: 1) the panel survey, which was addressed mainly to respondents in the 1991 survey, with certain additions; 2) the so-called supplementary survey, which focused on a new random sample of individuals; 3) the so-called nonresponse survey, which encompassed respondents who had participated in at least one of the earlier surveys but had since dropped out; 4) the time-use survey, which included the same sample of respondents as those in the panel and supplementary surveys. Individuals in the nonresponse group were not included in the time-use survey. Most of the questions in the first three surveys were the same, but certain questions sequences were targeted to the respondents in a specific survey. Thus certain retrospective questions were asked of the nonresponse group, while specific questions on social background, labor market experience...

The Household Market and Nonmarket Activities (HUS) project started as a joint research project between the Industrial Institute for Economic and Social Research (IUI) and Göteborg University in 1980. The ambition was to build a consistent longitudinal micro data base on the use of time, money and public services of households. The first main survey was carried out in 1984. In addition to a contact interview with the selected individuals, all designated individuals participated in a personal interview and two telephone interviews. All respondents were asked about their family background, education, marital status, labor market experience, and employment. In addition, questions about the household were asked of the head of household, concerning family composition, child care, health status, housing, possession of vacation homes, cars, boats and other consumption durables. At the end of the personal interview the household head had to fill out a questionnaire including questions about financing of current home, construction costs for building a house, house value and loans, imputation of property values and loans, additions/renovations 1983, maintenance and repairs, leasing, sale of previous home, assets and liabilities, and non-taxable benefits. All the respondents had to fill out a questionnaire including questions about tax-return information 1983, employment income, and taxes and support payments. Two telephone interviews were used primarily to collect data on the household´s time use and consumption expenditures. The 1986 HUS-survey included both a follow-up of the 1984 sample (panel study) and a supplementary sample. The 1986 sample included 1) all respondents participating in the 1984 survey, 2) the household heads, partners and third persons who should have participated in 1984 but did not (1984 nonresponse), 3) those individuals who started living together after the 1984 interview with an selected individual who participated or was supposed to participate in 1984, 4) members of the 1984 household born in 1966 or 1967. If entering a new household, for example because of leaving their parental home, the household head and his/her partner were also interviewed. Respondents participating in the 1984 survey were interviewed by telephone in 1986. Questions dealt with changes in family composition, housing, employment, wages and child care, and it was not only recorded whether a change had occurred, and what sort of change, but also when it occurred. The respondents also received a questionnaire by mail with questions mainly concerning income and assets. Respondents not participating in the earlier survey were interviewed in person and were asked approximately the same questions as in the 1984 personal interview. The 1988 HUS-survey was considerably smaller than the previous ones. It was addressed exclusively to participants in the 1986 survey, and consisted of a self-enumerated questionnaire with a nonrespondent follow-up by telephone. The questions dealt with changes in housing conditions, employment and household composition. The questionnaire also contained some questions on household income. In many respect the 1991 HUS-survey replicated the 1988 survey. The questions were basically the same in content and range, and the survey was conducted as a self-enamurated questionnaire sent out by mail. This time, however, in contrast to the 1988 survey, an attempt was made to include in the survey the new household members who had moved into sample households since 1986, as well as young people who turned 18 after the 1986 survey. Earlier respondents received a questionnaire by mail containing questions about their home, their primary occupation and weekly work hours since May 1988 (event-history data), earnings in 1989, 1990 and 1991, household composition and any changes in it that might have occurred since 1988, child care and some questions on income. New respondents were also asked about their education and labor-market experience. With respect to its design and question wording, the 1993 survey is a new version of the 1986 survey. The survey is made up of four parts: 1) the panel survey, which was addressed mainly to respondents in the 1991 survey, with certain additions; 2) the so-called supplementary survey, which focused on a new random sample of individuals; 3) the so-called nonresponse survey, which encompassed respondents who had participated in at least one of the earlier surveys but had since dropped out; 4) the time-use survey, which included the same sample of respondents as those in the panel and supplementary surveys. Individuals in the nonresponse group were not included in the time-use survey. Most of the questions in the first three surveys were the same, but certain questions sequences were targeted to the respondents in a specific survey. Thus certain retrospective questions were asked of the nonresponse group, while specific questions on social background, labor market experience...

The 2015 American Housing Survey marks the first release of a newly integrated national sample and independent metropolitan area samples. The 2015 release features many variable name revisions, as well as the integration of an AHS Codebook Interactive Tool available on the U.S. Census Bureau Web site. This data collection provides information on representative samples of each of the 15 largest metropolitan areas across the United States, which are also included in the integrated national sample (available as ICPSR 36801). The metropolitan area sample also features representative samples of 10 additional metropolitan areas that are not present in the national sample. The U.S. Department of Housing and Urban Development (HUD) and the U.S. Census Bureau intend to survey the 15 largest metropolitan areas once every 2 years. To ensure the sample was representative of all housing units within each metro area, the U.S. Census Bureau stratified all housing units into one of the following categories: (1) A HUD-assisted unit (as of 2013); (2) Trailer or mobile home; (3) Owner-occupied and one unit in structure; (4) Owner-occupied and two or more units in structure; (5) Renter-occupied and one unit in structure; (6) Renter-occupied and two or more units in structure; (7) Vacant and one unit in structure; (8) Vacant and two or more units in structure; and (9) Other units, such as houseboats and recreational vehicles. The data are presented in three separate parts: Part 1, Household Record (Main Record); Part 2, Person Record; and Part 3, Project Record. Household Record data includes questions about household occupancy and tenure, household exterior and interior structural features, household equipment and appliances, housing problems, housing costs, home improvement, neighborhood features, recent moving information, income, and basic demographic information. The Household Record data also features four rotating topical modules: Arts and Culture, Food Security, Housing Counseling, and Healthy Homes. Person Record data includes questions about personal disabilities, income, and basic demographic information. Finally, Project Record data includes questions about home improvement projects. Specific questions were asked about the types of projects, costs, funding sources, and year of completion.