Summaries of Articles

Abstract

Previous articleNext article FreeSummaries of ArticlesSummaries of ArticlesPDFPDF PLUSFull Text Add to favoritesDownload CitationTrack CitationsPermissionsReprints Share onFacebookTwitterLinked InRedditEmailQR Code SectionsMoreBuilding Fiscal Capacity in Developing Countries: Evidence on the Role of Information TechnologyMerima Ali, Abdulaziz B. Shifa, Abebe Shimeles, and Firew WoldeyesWeak state capacity is a major challenge in low-income countries, resulting in low revenue collection. As a result, these countries have relied on outside sources to finance development spending, often leading to a high risk of external debt distress and dependency on foreign aid. To overcome these challenges, many countries have begun implementing ambitious reforms to build the capacity of the state to increase tax compliance. This paper contributes to the limited but growing literature that examines the degree to which these reforms have been successful in achieving a high rate of tax compliance.In particular, the paper focused on evaluating the role of information technology in enhancing the capacity of the state to mobilize taxes, using unique administrative data from Ethiopia, which introduced electronic sales register machines (ESRMs) in 2008 and rolled them out in phases, covering nearly 70,000 firms in 2014. ESRMs offered the Ethiopian Revenue and Customs Authority (ERCA) unprecedented access to daily business transactions as the machines are linked with a central server maintained by ERCA. Given that indirect taxes constitute a high proportion of total revenue collected by the Ethiopian government, the potential of ESRMs in increasing tax revenue is high.However, the massive information generated by ESRMs may not necessarily translate into high tax revenue for various reasons, depending on the compliance rate by existing firms (intensive margin) and that of expanding the tax base (extensive margin). In addition, effectively using ESRMs for tax compliance requires a high degree of technical expertise, coordination among various government agencies, and reliable network infrastructure. The introduction of ESRMs also may increase or decrease the tax base depending on how an increase in tax enforcement affects overall demand, thereby encouraging firms to exit the formal sector. Or on the contrary, ESRMs may encourage firms to scale up operations as remaining small may no longer bring the expected benefit through tax evasion. The net effect of ESRMs hence depends on the tax administration system, behaviors of firms, and overall state capacity to effectively utilize the potentials of ESRMs.The study looked at the impacts of ESRMs on tax revenue through the lens of improved compliance (intensive margin) and expanding the tax base by encouraging entry and firm growth (extensive margin) using quasi-experimental methods. Results show that the adoption of ESRMs helped improve tax revenue through better reporting of total sales and value-added tax collections by registered firms (intensive margin). There was also evidence of declining firm entry following the adoption of the ESRMs, suggesting a decline in the tax base (the extensive margin). In addition, ESRMs are helpful only in monitoring total sales of firms but not the costs they incur (including intermediate inputs), which is important to increase tax compliance on profit taxes and reduce systematic tax frauds.The policy implication is that governments in developing countries may require a stronger political will and capacity to use third-party information by triangulating value chains across industries to harness the full benefits of information technology.The Distributional Effects of Property Tax Constraints on School DistrictsLucy Sorensen, Youngsung Kim, and Moontae HwangIn an effort to curb residents’ rising property tax burden, New York State in 2011 enacted a law that limits year-to-year growth in local property taxation to 2 percent or the rate of inflation, whichever is less. This limit applies to all local governments and school districts outside of Buffalo, Rochester, Syracuse, Yonkers, and New York City. Theoretically, such a policy should benefit property owners by keeping property tax rates in check. However, it could also harm the quality of public services if revenue losses force local governments or school districts to cut back on essential functions.The current study examines the impacts of the New York State property tax cap on school district finances and educational outcomes. It speaks to three primary questions. First, how much did the tax cap actually constrain school district property tax collection? Second, which types of districts, serving which types of students, did the tax cap affect most? And, third, did any revenue losses under the tax cap translate into changes in student performance?To conduct this study, we analyze over a decade of data drawn from the New York State Education Department, Office of the New York State Comptroller, and other publicly available data sources. To determine the level of constraint faced by each school district in each year, we compare the levies allowed under the cap to those we estimate that the district would have collected in the absence of the cap. This calculated value represents the degree of pressure exerted by the tax cap. We then estimate the effects of this pressure on school district revenues, expenditures, personnel, and student test scores.Our findings indicate that — even with a vote override provision built into the law — the tax cap significantly reduced school districts’ local revenue collection. We estimate that districts across New York State lost a total of $8.4 billion in education revenues by 2016. The cap affected wealthier school districts more so than economically disadvantaged school districts, because wealthier districts rely more heavily on property taxes to fund their schools. School districts facing higher levels of revenue constraint also experienced substantial losses in student reading and math performance in the years following the tax cap. This appears to be explained by reductions in instructional expenditures, teacher hiring, and support personnel.Our study demonstrates that tax and expenditure limitations such as that of New York State not only rein in government spending but also directly harm the quality of public services — in this case, public education. Future research should investigate how tax cap policies interact with other systems such as state school finance equalization reforms or federal education aid packages.Generosity Across The Income And Wealth DistributionsJonathan Meer and Benjamin A. PridayThe relationship between income, wealth, and charitable giving is of academic and public interest. Despite the topic’s importance, there has been little systematic evidence to robustly understand the relationship. The existing literature — drawing on data from the United States and several European countries — has largely focused on the “U-shaped giving curve”: households at the bottom of the income distribution give the largest proportion of their income, high-income households give the second largest, and middle-income households give the least. Previous estimates of this relationship have suffered from a lack of available data, misspecification, and omission of important covariates.We use nine waves of the Panel Study of Income Dynamics (PSID), spanning 2001–2017, to provide a more robust understanding of the income-wealth-giving gradient across three measures of household “generosity”: the likelihood of making a donation, the amount given, and the proportion of income given. Our data provide important, distinct advantages over existing estimates. The PSID’s measurement of giving is comparable in accuracy to federal tax return data but includes low-income households that are omitted from tax data because they do not itemize deductions. The panel nature of the PSID allows us to econometrically remove the confounding effect of any time-consistent, household-level differences in unobserved altruistic preferences that affect giving. The PSID also contains measurements of wealth, which, as we show, is an important factor when estimating how household resources are related to giving; omitting wealth from analysis misrepresents the relationship between income and giving.We report simple means, as well as ordinary least squares and fixed-effects estimates, of three observable measures of generosity. We find that, irrespective of specification, the propensity to donate money and the amount given increase with a household’s resources. In contrast to much of the existing literature, we show that the mean percentage of income given is relatively flat across the income distribution after accounting for a small number of extreme observations. These “outlier” households are primarily counted as low income but tend to be older and more educated and have higher wealth. As such, they are not truly indicative of low-income households, and including them in simple means misstates the income-giving gradient. The PSID also disaggregates giving by causes: 10 secular categories (e.g., health, humanitarian aid, environment) and religious giving. We show, consistent with existing empirical research, that as income increases, households allocate a smaller proportion of their giving to religious services. However, even the highest-income group in our sample (mean annual income of $414,400) gives more to religious services than any other cause.Leakage from Retirement Savings Accounts in the UNITED STATESLucas Goodman, Jacob Mortenson, Kathleen Mackie, and Heidi R. SchrammEmployer-sponsored defined contribution (DC) retirement plans and individual retirement arrangements (IRAs) are important vehicles for retirement savings for many US households. The federal government subsidizes these plans heavily, forgoing roughly $250 billion in tax revenue in 2019, with the primary goal of encouraging retirement adequacy for middle-class individuals.One particular concern about these plans is that some savings may “leak” out of the accounts prior to retirement, potentially reducing retirement savings. Yet the socially optimal amount of leakage is unlikely to be zero. On the one hand, the ability to draw down assets prior to retirement allows people to smooth their consumption around income shocks such as unemployment. This flexibility likely encourages saving in these accounts. On the other hand, it is well known that people suffer behavioral biases in the context of retirement saving, such as present bias, meaning that observed withdrawals might not necessarily make people better off. Additionally, there are substantial transaction costs that savers must incur in order to transfer their funds when they change jobs, and people may be choosing to cash out their savings primarily to avoid these costs.This study makes two main contributions. First, we provide improved estimates of preretirement distributions to individuals. We use a large panel of administrative tax data and carefully use self-reported and third-party information to identify and classify distributions. Second, we use the individual-level panel data to examine correlations between leakage, individual characteristics, and common preretirement-age life events. We provide a detailed analysis of leakage that occurs in the same year as a job separation, which is the strongest predictor of leakage in our data.We estimate that from 2003 to 2015, total DC and IRA withdrawals by preretirement-age individuals were equal to 22 percent of contributions by this age group. Consistent with consumption smoothing motives, income is strongly correlated with leakage: those in the lowest income decile leak nearly 10 times more (as a share of contributions) than those in the top income decile. The distribution of leakage across individuals is also highly skewed. Among those who made a DC or IRA contribution between 2003 and 2015, over half did not make a withdrawal, while about 25 percent withdrew more than their total measured contributions.We next analyze how leakage evolves around certain life events. Most notably, we find that job separation is associated with a near tripling of the probability of leakage. We further analyze the extent to which this association might be driven by consumption smoothing motives versus transaction costs. We find that leakage is substantially higher for those job separators who experience a reduction in income at job separation, though leakage remains elevated even for those that experienced an increase in income (i.e., due to moving to a higher-paying job). This suggests that both consumption smoothing and transaction costs are playing a role in driving leakage at job separation.Estimating the Distributional Implications of the Tax Cuts and Jobs ActCody Kallen and Aparna MathurThis paper provides a distributional analysis of the Tax Cuts and Jobs Act (TCJA) that became law in December 2017, with implications for American households from changes to the individual income tax as well as changes to the corporate income tax. We model the impact on average tax rates, after-tax incomes, and other distributional measures, using tax records data from the 2011 Internal Revenue Service public use file and the open-source Tax-Calculator microsimulation model.When appropriately accounting for differences in tax unit size, through 2025 the individual income tax changes of the TCJA appear progressive. Failing to account for these differences obscures these progressive effects because larger families benefiting from the expansion of the child tax credit have higher incomes. Combining individual income and business tax provisions in the static scenario, the TCJA is likely to result in short-term increases in posttax incomes across all deciles.If the individual income tax provisions are allowed to expire, by 2027 the TCJA becomes regressive, with most filers facing higher taxes but high-income filers receiving higher after-tax incomes due to the remaining corporate tax cuts. Long-run dynamic effects may increase after-tax incomes on average across all income groups or at least partially offset the tax hikes, but the effects are regressive, with income gains skewed toward the top.In addition to the distributional results, this paper highlights important methodological issues when conducting a distributional analysis. We identify the importance of choosing appropriate equivalence scales to compare incomes across different types of tax filers or households, with nontrivial implications for measuring progressivity or regressivity. The relevance of equivalence scales to the distributional analysis of the TCJA extends to any tax reform with heterogeneous effects across households of different sizes. Modeling both the static as well as the dynamic distributional impact, we find that the effects of the TCJA vary significantly across and within income deciles, depending upon family structure, the choice of equivalence scales, and the year of analysis. Previous articleNext article DetailsFiguresReferencesCited by National Tax Journal Volume 74, Number 3September 2021 Published for: The National Tax Association Article DOIhttps://doi.org/10.1086/717535 Views: 185 © 2021 National Tax Association. All rights reserved.PDF download Crossref reports no articles citing this article.