{"paper_id":"2408.14476","title":"A new approach to the theory of optimal income tax","authors":["Vassili N. Kolokoltsov","Egor M. Dranov","Denis E. Piskun"],"paper_date":"2024-08-10","paper_type":"theoretical","paper_url":"https://arxiv.org/pdf/2408.14476v1","page_url":"https://thequant.space/flowcharts/a-new-approach-to-the-theory-of-optimal-income-tax/","scores":{"empirical_rigor":1,"math_complexity":8,"quadrant":"Lab Rats"},"proposes":{"status":"available","statement":"The paper proposes a new two‑parameter paradigm for optimal income taxation that maximizes total utility while also considering the standard deviation of utility.","quote":"The second objective is to suggest a new paradigm for optimal taxation, where instead of just total average utility maximization one introduces a standard deviation of utility as a second parameter (in some analogy with Marcowitz portfolio optimization).","summary":"The Nobel-price winning Mirrlees’ theory of optimal taxation inspired a long sequence of research on its refinement and enhancement. However, an issue of concern has been always the fact that, as was shown in many publications, the optimal schedule in Mirrlees’ paradigm of maximising the total…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Individuals have utility u(c,l)=c−l^2/2 (simplest natural utility function).","type":"model","stated":"explicit","quote":"In Section 2 we prove the result that, under the simplest personal utility function u(c, l) = c−l2/2 and linear or piecewise linear (with one kink) tax function, the optimal taxation in the standard setting of maximal total utility is simply no taxation at all.","support":"full"},{"assumption":"The tax function t(y) is continuous and nondecreasing.","type":"model","stated":"explicit","quote":"Mechanism design of a government is specified by a tax function t(y), which is assumed to be continuous and nondecreasing, so that the income of an individual after tax is c(nl) = y−t(y) = nl−t(nl).","support":"full"},{"assumption":"The distribution of productivity parameter n, f(n), is exogenously given.","type":"data","stated":"explicit","quote":"2 Piecewise-linear tax in the standard setting Recall that in the standard approach to optimal income tax (see [5] and [9]), it is assumed that individuals are characterised by a parameter n ≥0, yielding their production per unit of efforts (e.g. time), with a (exogenously given) distribution f(n).","support":"full"},{"assumption":"The tax revenue is fully redistributed (budget constraint ∫ t(y)f(n)dn = 0).","type":"model","stated":"explicit","quote":"(1) Since lmax are not unique only on a discrete set of n, this non-uniqueness does not affect the value of U. We also assume, as in [9], that the tax is fully distributed, so that Z ∞ 0 t(y)f(n)dn = Z ∞ 0 t(nlmax(n))f(n)dn = 0.","support":"full"},{"assumption":"The analysis is restricted to piecewise‑linear tax environments with at most one kink.","type":"model","stated":"explicit","quote":"We shall work with the piecewise-linear tax environment suggested in [10] and further analysed e.g. in [1], [11], [2].","support":"full"},{"assumption":"Linear tax case is considered as a special case of the piecewise‑linear framework.","type":"model","stated":"explicit","quote":"In Section 2 we prove the result that, under the simplest personal utility function u(c, l) = c−l2/2 and linear or piecewise linear (with one kink) tax function, the optimal taxation in the standard setting of maximal total utility is simply no taxation at all.","support":"full"}]},"datasets":{"status":"none_found","items":[],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"none"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"In the standard setting with the simplest utility, the optimal taxation is no taxation at all.","kind":"theoretical","quote":"In Section 2 we prove the result that, under the simplest personal utility function u(c, l) = c−l2/2 and linear or piecewise linear (with one kink) tax function, the optimal taxation in the standard setting of maximal total utility is simply no taxation at all.","support":"full"},{"finding":"The two‑parameter paradigm yields natural optimal linear tax rates that depend on a parameter balancing total utility growth against utility standard deviation.","kind":"theoretical","quote":"The result is a very natural (and easy to interpret) optimal linear taxes depending on the choice of a parameter that quantitatively assesses the acceptable ratio between the growth of total utility and the growth of its standard deviation, the latter considered as a measure of social tension in a society.","support":"full"},{"finding":"Analogous results hold for the piecewise‑linear two‑bracket case, with numerical analysis confirming the same qualitative conclusions.","kind":"theoretical","quote":"Here analytical solutions are much more difficult to obtain, but numeric analysis is not too complicated to perform, leading essentially to analogous results, as in the linear case.","support":"full"},{"finding":"The proposed framework also works with a logarithmic personal utility function.","kind":"theoretical","quote":"In Section 5 we show how our program of two-criteria optimal taxation works with another example of personal utility, the logarithmic one.","support":"full"},{"finding":"The optimal β (tax progressivity) can take any value in [0,1] depending on the chosen trade‑off parameter c.","kind":"theoretical","quote":"Thus we see that the optimal β(c) can take any value from [0, 1] depending on the choice of c.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":17,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":2,"extracted_at":"2026-11-14T19:56:05Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}