10 - Tax enforcement Flashcards
what is the relationship between reported income and
p = probability of audit
pheta = fine
w- increases with both
- so evasion will be less - because reporting higher income
why is observed tax evasion levels much lower than the model works out
- when we add in the values for p and pheta
- unwilling to cheat
- unable to cheat
what is unwilling to cheat
- social norms
- morality
- dont want to cheat
what is unable to cheat
- may want to cheat
- but cant because of institutional features
- 3rd party reporting - when employer is reporting on your behalf
what are the 3 kinds of experimental data that can be used to measure evasion
- since not observable
- IRS
- lab experiments
- field experiments
what do IRS studies do
- how do they measure evasion
- carries out random audits to estimate the tax gap
- checks all their documentation
- estiamates for population
what is the tax gap
taxes evaded/taxes owed
what should be paid - what is paid = taxes evaded
what do IRS find from their audit checks
- US tax gap = 14%
- tax gap concentrated among self-reporting income
what is the UK tax gap like
HMRC
- tax gap is declining - 5.1%
- 1% through PAYE (third party)
- self assessment has more scope for non-compliance
what is the share of tax gap coming from each source in UK
- highest share of tax evasion - in personal taxes = but brings in low revenue
- second highest share = VAT misreporting = and brings in high revenue
- within personal taxes - basically all missing revenue is coming from self reported not PAYE
so what is the area that UK should target tax enforcement
VAT misreporting
and top earners
- out of self employed that are misreporting the top 4% account for 42% of lost revenue
what are ways that HMRC can increase tax compliance
- audit (costly)
- direct reporting
- third party reporting
- behavioural interventions
Advani (2021)
what do they test
what do they do
effects of UK audits on LR compliance behaviour
- does randomly audited a group have any LR effects in the amount they report in the future
- is there a benefit of auditing
- use admin databases on UK taxes
- randomised audit program
- Treatment = given audit
- Control = no audit
Advani (2021)
what do they find
- before audit = T and C no difference in reportings
- after audit = T group tax compliance is higher than control - reports higher incomes for 5 years
- audits constrain future misreporting
- audits can be used as threats
- audits have long lasting effects
why is measuring tax evasion through lab experiments bad
- better to look at field experiments
- lab experiments overestimate tax evasion
- they do find strong relationship between p and pheta and reduced evasion
- missing real world aspects
- games with students arent the same (social norms + 3rd party reporting)
Blumenthal (2001)
what do they do
- use normative appeals to comply
- T = send non-threating letters to encourage compliance
- c = no letter
Blumenthal (2001)
what do they find
- no statistically significant effect of normative appeals on compliance overall
- these letters dont impact tax compliance
Slemrod (2001)
what do they do
- T = send threat of audit letter
- scare people into compliance
- C = no letter
Slemrod (2001)
what do they find
- downfall of paper?
- statistically significant effect on reported income increase
- especially among self employed
- threats = work
- normative = dont work
- but small sample
what are key descriptive data from Denmark
- among evaders the most common strategy is to misreport everything
evasion by fraction of income self reported
- 3rd party evasion rate = 0
- evasion rate is 30% for positive fraction of self reported income
- % of income evaded is increasing in the share of self reported income
- below 0.2, misreport all income
- above 0.2, the percentage of income misreported grows slower = get scared
Kleven (2011)
what did they do
- Denmark study of 40,000 sample
- split into 2 groups - 100% audited in 2007 and not audited
- fiscal year after split into 3 groups - no letter, 100% you will be audited letter, 50% you will be audited
- followed up with audit in 2008?
kleven what did they find
stats
- self employed evasion rate is almost 40%
- third party evasion rate is 0.3%
- overall evasion rate is 2.5%
- because 95% of income is 3rd party reported in denmark (unable to cheat)
what are klevens results from
marginal tax rates
- bunching before and after audit
- 100% audit group in 2007, before is 2006
- find that most bunching is not due to evasion but avoidance
- effect of MTR of evasion is small
- after audit - bunching only decreased by 1/3
- only 1/3 of people started to report honestly and stopped bunching
- means 2/3 is explained by something else
- no change in bunching in stock income = means that all that bunching is driven by avoidance - income shifting, intertemporal
what is kleven result from
prior audit effects
- significant effects on reported income increases for 100% audited group compared to 0%
- driven by self reported items
what is kleven result on
threat of audit
- sigificant effects on self reported income increases
- more increase for people that got the letter then didnt
why does the AS model fail
by overestimating tax evasion
- p is very low in real life - and still see low evasion
- doesnt take into account third party reporting
- most income is reported this way and it is very hard to evade
- unable to evade
what happens when elasticity of the detection probability wrt undeclared income is
e = 0
- then always evade if p(1+pheta)<1
what happens when elasticity of the detection probability wrt undeclared income is
e > 0
- evading more increases risk of being caught on all evaded taxes
what is the shape of the
p(w-) function
s shape
- probability jumps to 1 when income starts to include 3rd party reported
- probability is low for all self reported
what is
p(w-)
- detection probability as a function of tax evasion depending on the share of self reported income
so how was the simple AS model wrong
- was assuming probability of getting caught was constant
- p = 1/1+pheta
- p is a function of w-
- p = 1/(1+pheta)(1+e)
- the optimum is just to the left of w-s
why is it hard for evasion to happen in third party reporting
- single employee can unfold the collusion
- hard to mess with accounting and payroll records
what is
turnover taxes
sales taxes
VAT
- which is easier to enforce
- tax all sales - B-B and B-C
- B-C sales only - strong incentive of evading
- B-B and B-C - only on value added = easy to enforce
what is the automatic enforcement mechanism of VAT
- since you only pay taxes on value added (sales - purchases)
- incentive to make sure the B you buy your inputs from declares those taxes so that you can deduct the inputs from your taxes
- B-B wants VAT receipt
- strong incentive to pay taxes on inputs
Pomeranz (2015)
what do they do
- randomised experiment in Chile
- send VAT threat letters to businesses
- T = get letters
- C = no letters
Pomeranz (2015)
what do they find
- threat letters increase VAT collected - significant
- increase is driven mainly by final sales compliance = consumers dont have incentive to not evade
- small impact on intermediate sales = already reporting VAT fully
- evidence that VAT is self enforcing between B-B, but not B-C
Naritomi (2019)
what did they do
what did they want to find
- will giving consumers incentive to make sure firms report final sales increase tax revenue
- give consumers tax rebates and monthly lottery prizes for consumers that ask for receipts
- consumers upload the receipt
- treatment = retail sales
- control = wholesale sector - no effect on lottery because already reporting VAT fully
Naritomi 2019
what did they find
- giving consumers incentive to ensure firms report final sales transactions
increased tax revenue by 9.3%
- but this is a costly way to enforce VAT so not good for LR
how big of an issue is holding wealth offshore
- 8% of global financial wealth of hhs is held in tax havens
- probability to own an unreported HSBC account increasese by wealth group
- a lot of missing revenue to gov - because super wealthy are hiding money in tax havens the most
what are methods to reduce curbing offshore tax evasion
- require banks to report accounts owned by US persons or face stiff penalties
