-
Notifications
You must be signed in to change notification settings - Fork 0
/
research.html
788 lines (588 loc) · 47.8 KB
/
research.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
<!DOCTYPE html>
<html lang="en">
<head>
<title>Adrien Pacifico </title>
<!-- Global site tag (gtag.js) - Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=UA-124436397-1"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'UA-124436397-1');
</script>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/css/bootstrap.min.css">
<link rel="stylesheet" href="https://use.fontawesome.com/releases/v5.2.0/css/all.css" integrity="sha384-hWVjflwFxL6sNzntih27bfxkr27PmbbK/iSvJ+a4+0owXq79v+lsFkW54bOGbiDQ" crossorigin="anonymous">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.3.1/jquery.min.js"></script>
<script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js"></script>
<script type="text/javascript" charset="UTF-8" src="./Bilan démographique 2018 - Insee Première - 1730_files/graphsvg.js"></script>
<script>"https://cdnjs.cloudflare.com/ajax/libs/jquery.tocify/1.9.0/javascripts/jquery.tocify.min.js"</script>
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/jquery.tocify/1.9.0/stylesheets/jquery.tocify.min.css">
<script type="text/javascript" async src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/latest.js?config=TeX-MML-AM_CHTML"></script>
<style>
body {
position: relative;
}
div.col-sm-9 div {
height: 250px;
font-size: 28px;
}
.affix {
top: 200px;
z-index: 9999 !important;
}
hr {
display: block;
margin-top: 0.5em;
margin-bottom: 0.5em;
margin-left: auto;
margin-right: auto;
border-style: inset;
border-width: 2px;
}
</style>
<body data-spy="scroll" data-target="#myScrollspy" data-offset="20">
<div class="jumbotron text-center" style="margin-bottom:0">
<h1>Adrien Pacifico, Ph.D.</h1>
<p>Economics of Taxation, Fiscal Microsimulation, Data Science </p>
</div>
<nav class="navbar navbar-inverse">
<div class="container-fluid">
<div class="navbar-header">
<button type="button" class="navbar-toggle" data-toggle="collapse" data-target="#myNavbar">
<span class="icon-bar"></span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
</button>
<a class="navbar-brand" href="#"></a>
</div>
<div class="collapse navbar-collapse" id="myNavbar">
<ul class="nav navbar-nav">
<li><a href="index.html">About Me</a></li>
<li class="active"><a href="./research.html">Research</a></li>
<li><a href="./policy_briefs.html">Policy Briefs</a></li>
<li><a href="./other.html">Other Work</a></li>
</ul>
</div>
</div>
</nav>
<div class="container">
<div class="row">
<nav class="col-sm-4" id="myScrollspy">
<ul class="nav nav-pills nav-stacked hidden-xs" data-spy="affix" data-offset-top="205">
<h3>Working-papers: </h3>
<li class="active"><a href="#lowering">The lowering of the family quotient</a></li>
<li><a href="#cohabitants">Cohabiting couples</a></li>
<li><a href="#tax_frequency">Tax frequency</a></li>
<li><a href="#future_research">Future Research</a></li>
</ul>
<hr class="hidden-sm hidden-md hidden-lg">
</nav>
<div class="col-sm-8">
<div id="lowering">
<h2>
The lowering of the family quotient
</h2>
(Job Market Paper)<br>
<a href="./docs/bokeh_graph.html">Graph Full Screen</a>,
<a href="https://mybinder.org/v2/gh/adrienpacifico/adrienpacifico.github.io/master?filepath=Notebooks%2Fbokeh_graph_for_website_qf_lowering.ipynb" target="_blank">Generate Graphs and Thresholds with OpenFisca,</a><a href="./docs/slides/slides_qf_lowering/index.html"> Slides.</a>
<h5>Research question: </h5>
We want to know how rich households reacted, on their taxable income margin, to a tax reform that induced, for some households a lump sum increase in their tax liability, or a lump sum increase and a change in marginal tax rate for other households.<br>
I rely on the French joint income tax that applies different schemes based on family composition as an additional source of variation. It allows me to do a triple difference estimation of the reform.
<div class="embed-responsive embed-responsive-16by9">
<iframe class="embed-responsive-item" src="./docs/bokeh_graph.html" >
</iframe>
</div>
<script src="./bokeh_graph.js" type="text/javascript"></script>
<div class="bk-root" id="21339a25-362e-4b60-bb33-a2dd1c043c02"></div>
<h5>Context: </h5>
<ul>
<li>The French tax system works such that the <i>child tax break </i> is increasing with income.
</li>
<li>
However, the fiscal gain a household can derive from a child is bounded by a ceiling.
</li>
<li>
The ceiling per child has been lowered from 2336€ to 1500€ between 2011 and 2014. That per child ceiling is twice the baseline amount for the third child and over.
</li>
<li>
Households have been affected in three ways:
<ol>
<li>
Households for which the fiscal advantage per child was lower than 1500 euros are unaffected.
</li>
<li>
Households for which the fiscal advantage per child was between 1500 euros and 2336 euros face a change in their marginal tax rate.
</li>
<li>
Households for which the fiscal advantage per child was over 2336 euros face a lump sum loss in disposable income of 836€ per child.
</li>
</ol>
</li>
</ul>
<h5>The idea: </h5>
<ol>
<li>
Households below the ceiling are not impacted by the reform and can be a control group.
</li>
<li>
Households between 2011 and 2014 ceilings will be subject to an <b>income and a substitution effect</b>.
</li>
<li>
Households that are over the ceiling, since they incur a lump sum loss (without any change of their marginal tax rate) are thus subject to a <b>pure income effect</b>.
</li>
</ol>
<h5>Implications: </h5>
As the fiscal advantage increases with income, there exists an income threshold at which a household will attain the ceiling. <br>
The graph above shows the fiscal tax break due to children for a married couple from one to three children before and after the reform.
The following table gives the taxable wage for a given number of children at which ceiling is attained.
<h6>
<div >
<table border="3" class="table table-hover table-sm table-condensed" >
<thead>
<tr style="text-align: right;" width=30%>
<th></th>
<th>2011</th>
<th>2013</th>
</tr>
<tr>
<th>Children #</th>
<th></th>
<th></th>
</tr>
</thead>
<tbody>
<tr>
<th>1</th>
<td>67776</td>
<td>62624</td>
</tr>
<tr>
<th>2</th>
<td>79000</td>
<td>68028</td>
</tr>
<tr>
<th>3</th>
<td>101391</td>
<td>79238</td>
</tr>
<tr>
<th>4</th>
<td>123866</td>
<td>90302</td>
</tr>
<tr>
<th>5</th>
<td>145307</td>
<td>101424</td>
</tr>
<tr>
<th>6</th>
<td>159450</td>
<td>112324</td>
</tr>
</tbody>
</table>
</div>
</h6>
<h5>Data: </h5>
I use the Échantillon Démographique Permanent (EDP). The EDP allows to follow a panel of 2 million households over 5 years with their tax returns. It also contains a rich set of other variables such as diplomas, housing characteristics, that allows to include a rich set of controls to the analysis.
<h5>The method: </h5>
As emphasized by <a href= "http://piketty.pse.ens.fr/fichiers/public/Piketty1999f.pdf" data-toggle="tooltip" data-placement="right" title="L'immense avantage du mécanisme français du QF,
unique en son genre parmi les grands pays
développés, est qu'il permet d'identifier des
''groupes de contrôle'' relativement fiables, sous la
forme de groupes de contribuables dont le niveau de
revenu imposable est exactement le même que ceux
du ''groupe test'', mais qui ne sont pas,concernés par
les mêmes variations de taux marginaux
d'imposition, du fait d'un nombre de parts de QF
différent de celui du ''groupe test''. ">Piketty(1998)</a>, the family quotient mechanism embodies in itself a <u>triple difference estimator</u>. As the condition to be treated is conditionnal to be on a specific position of the income distribution for a given number of children, we can run a triple difference estimator:
<p>$$\begin{align*}
\Delta y_i &= \beta_0 + \sum_{i=1}^{6} {\beta_i \text{Children}_i} + \sum_{j=1}^{6} {{}_{b}\delta_i \text{Between}_i} + \sum_{i=1}^{6} {{}_{o}\delta_i \text{Over}_j} \\
&+ \sum_{i=1}^6 {{}_{b}\gamma_{i} \text{Children}_i \times \text{Between}_i} \\
&+ \sum_{i=1}^6 {{}_{o}\gamma_{i} \text{Children}_i \times \text{Over}_i} \\
& + \epsilon_i
\end{align*}$$ </p>
<p> \(\Delta y_i \), the endogenous variable, is the change in taxable income before the reform (in 2011) and after the reform (in 2014).
</p>
<p> \(Over_i\) is the dummy indicator for being over the income threshold where the fiscal advantage is saturating the 2011 ceiling constraint per i children (e.g 6777 euros for 1 child families). \(Between_i\) the one for being between 2011 and 2013 thresholds. \(\text{Children}_i\) the dummy indicator for i number of children. </p>
<p>Under the classical assumption underlying the classical diff-in-diff literature, \(\beta_0\) captures the general trend in the population, \( \beta_i \) the average difference in trend based on the number of children in the household, \( {}_{b}\delta_i \) and \( {}_{o}\delta_i \) captures the effect of being between or over the tresholds. </p>
<h5>Results: </h5>
<p> We find evidence of large income effect on that population, contrary to cross-sectional analysis of elasticities, or to income tax reforms that concern the top 1% of the income distribution. It is very likely that these large income effects are due to habits formation, meaning that households do not accept to diminish their standards of living by compensating the income loss associated with the tax reform by working more. Those results are preliminary.</p>
<hr>
<br>
</div>
<div id="cohabitants">
<h2>A Direct Test of (In)efficiency within Couples:<br> Tax Optimization in French cohabiting couples.
<a href="#cohabiting"></a>
</h2>
(with O. Bargain, D. Echevin & N. Moreau)<br>
<a href="./Notebooks/cohabitants-heatmap.html">Graph Full Screen</a>,
<a href="https://mybinder.org/v2/gh/adrienpacifico/adrienpacifico.github.io/master?filepath=Notebooks%2Fplotly_openfisca_cohabitants.ipynb" target="_blank">Generate Graphs with OpenFisca.</a>
<h5>Research question: </h5>
We want to test the Pareto efficiency assumption made in collective households models by relying on a natural experiment that concerns cohabiting couples with children. Those couples have to choose to which tax-unit to allocate their children. Some allocation leads to a higher income tax than others, and households that do not select the allocation that leads to the smallest income tax are not respecting the Pareto efficiency assumption. We document the number of households that do not have efficient behaviors, then we try to investigate sources of these inefficiencies by investigating different pathways.
<div class="embed-responsive embed-responsive-4by3">
<iframe class="embed-responsive-item" src="./Notebooks/cohabitants-heatmap.html"></iframe>
</div>
<p>
<b> The setting: </b>
<ul>
<li>Cohabitants living in France under the same household constitute two fiscal units.
</li>
<li>In the presence of children in the household, they can allocate each non-stepchild to one of the two tax units.<br>
For example, a family of two cohabitants and two children, there exists three allocations: both children on the mother tax return, one child on each cohabitant tax return, and both children on the father tax return.
</li>
<li>
Putting a child on a tax return leads to a tax break that is dependent of the taxpayer income.
Depending on the mother and father income, each allocation will lead to a defined income tax liability.
</li>
</ul>
The graph above, made to represent a family with two children, provide for each cohabitant income combination and each allocation, the associated income tax for the year 2013. The color grading represents the difference between the best and the worst allocation. Allocation A corresponds to all the children put on the father, B the equal split, and C all on the mother.
</p>
<p>
<b> The idea: </b>
The assumption usually made over households is that they respect efficiency. Cohabitants with children should thus choose the allocation that minimizes their tax liability. The choice of allocation is a quasi-experimental framework that allows doing a simple test of efficiency within the household.
</p>
<p>
<b> Results: </b>
We use a big administrative dataset (EDP) which allows us to compute the income tax of each cohabitant, but also eliminate stepfamilies from the sample (with civil registry databases). After computing the tax of each allocation an household can have, we compare the allocation leading to the smallest tax liability to the chosen one. We show that over 25% of the households do not respect the Pareto efficiency by choosing an allocation that is not optimal.</p>
<p>
We then investigate several avenues to explain why some couples are inefficient.
<ul>
<p>
<li> <b>Social Norm</b>: We show that couples are biased, as a part of the allocation errors result to the fact that households tend to allocate more children to the father.</li>
</p>
<p>
<li> <b>Inertia</b>: While 20% of the households change their allocation from one year to another to minimize their tax liability, 80% of the households that should have changed their allocation did not by sticking to their previous year allocation. About 8% of the households "learn" the tax scheme by changing their allocation for the optimal one. Pure confusion (or erratic choice) only account for about 2% of the non-optimisation behavior. </li>
</p>
<li> <p> <b>Cooperation</b>: Couples to optimize need to cooperate. They have to share their exact income, compute the income tax of each allocation, then probably bargain for a cash transfer from the cohabitant taking the most children to the one taking the less.</p>
<p> We make the assumption that cooperating couples tend to marry or engage in a civil union more, and that non-cooperating couples tend to separate. </p>
<p> We then test the cooperation assumption by running a multinomial logit regression on the change in the marital status over whether or not the allocation chosen is the optimal one. We show that not optimizing with an income loss that is greater than 1% of the household income implies a greater probability to separate while optimizing increase the probability to get married or to contract a civil union (PACS). </p>
</li>
</ul>
</p>
</div>
<hr>
<div id=tax_frequency>
<h2>Tax frequency
<a href="#tax_frequency"></a>
</h2>
(with O. Bargain and Alain Trannoy)
<h5>Research question: </h5>
<p>
In this article, we explore the temporality of the tax by comparing a tax system with an income tax computed on an annual basis but paid monthly (one-twelfth each month), to a tax computed on a monthly basis and paid monthly.
</p>
<img class="img-responsive" src="docs/tax_frequency_graph.png" alt="graph tax-frequency" id="image_me">
<p>Formally we compare:
<ul compact>
<li>the first system where the tax paid at a given month \(t \) is \(\frac{G(\sum_{1}^{12}y_{t})}{12}\),
</li>
<li>to one where the tax paid at a given month \(t \) is \( \frac{G(12 y_{t})}{12} \).
</li>
</ul>
With \(G()\) a general tax function, an \(y_t\) the tax-unit income for a given month \(t\).
</p>
<p>The question of the frequency of the tax is of little interest if streams of income are steady. The monthly tax streams would be the same between the two tax systems. <br>
However, if they vary, in the presence of a convex tax two opposite force will be at stake:
<ul compact>
<li><b>A higher tax effect:</b> <br>
If a tax is convex, due to Jensen's inequality, agents with varying income will have a higher tax liability with an annual scheme than with a monthly scheme.
</li>
<li><b>An income smoothing effect:</b> <br>
A concave utility implies that agents want to smooth their consumption between periods. If a tax is monotonously increasing, a monthly tax will have an insurance/income smoothing property that will make agents better off. In other words, people prefer to pay a higher tax when they earn more, and a lower tax when they earn less than to pay always the same tax whatever their income. The "more convex" the tax function is, the higher the income smoothing effect is.
</li>
</ul>
<p>
We then run a simulation over the French income tax with monthly data (French Labor Survey) to see which effect dominates. To do so we take an HP Young equal sacrifice utility function derived from the income tax scheme with an inter-temporally separable utility function.
We find that the monthly scheme is in aggregate better than the annual scheme with an aggregate money metric gain of 2 billion euros.<br>
The effect is mainly driven by the bottom of the income distribution for three reasons:<br>
1) Households at the bottom of the income distribution are the ones that face the most yearly variations in income. 2) These households with the lowest income are the ones that are on the steepest part of the utility function and thus weight more in the utilitarian appraisal of the monthly scheme. 3) Income tax brackets are very large and not numerous, thus most households that have small variations of income will face a linear tax scheme, and thus only the income smoothing effect will apply.
</p>
<p>
We then test two other temporal tax system:
<ul compact>
<li><b> An equivalized tax scheme</b> that is a monthly tax scheme for which the tax liability is the same than with the annual tax scheme.
$$\sum_{1}^{T}\frac{1}{T+\lambda}G(Ty_{t})$$
</li>
<li><b>And an income averaging tax system</b> that follows the tax system proposed by <a href="https://www.journals.uchicago.edu/doi/abs/10.1086/255390?journalCode=jpe">Vickrey (1939)</a> which aims at the tax to be completly neutral (in the amount of money) with respect to the time at which an income is realized.<br>
$$ {g_{t}^{V}}= \frac{t}{T} G\left(T\frac{\sum_{p=0}^{t}y_{p}}{t}\right) -\sum_{p=0}^{t-1}g_{p}^{V}$$
</li>
</ul>
We show theoretically that both of these tax systems are Pareto improving with respect to the annual and monthly tax system. For increasing income streams the equivalized tax scheme dominates the income averaging scheme, and for decreasing income, the income averaging scheme dominates the equivalized scheme.
</p>
<p>
With our previous simulations assumptions, we run simulations over these two new income tax scheme. The compensated scheme implies a gain of 500 million euros over the monthly scheme, while the averaging scheme implies a gain of 3 billion euros over the compensated scheme.
<!--
</p>
=======
</p>
- Put reproducible science in the "other" section
<b>Conclusion:</b>
In this article, we give insights on the temporality of the tax which has been relatively unexplored since Vickrey. <br>
It sheds light
We have theoretically, given a static tax scheme, compared several temporal implementations of that scheme. We then have run simulations on the French tax system which exhibit that substantial welfare gains can be obtained compared to the actual tax scheme. Over those tested solutions results favors the proposition that Vickrey made in 19xx. <br>
We also show that the subject is mainly at stake for the bottom of the distribution. Indeed <br>
Our research is also quite
</p>
<p>
<b>Interpretation and Policy Implications:</b>
Vickrey larger demogrant smaller gains.
Vickrey larger demogrant ==> smaller gains.
>>>>>>> - Put reproducible science in the "other" section
Equivalized, not implementable.
Parfit vs. Vickrey
The domination of the Vickrey scheme (that implies negative income tax if present income is small enough compared to previous periods average income) is partly due to the chosen utility function: H.P Young utility functions asymptotically tend towards \(-\infty\) for a null income. A consequence is that a demogrant would reduce the money metric gains between the compensated and the averaging scheme. <br>
However, the compensated scheme lack of an implementation problem: it implies to know future realized income before applying the tax scheme to a specific tax unit.
</p>
<p>To answer that question we are making the assumption that the yearly utility function is inter-temporally additive monthly function increasing in income \(z\): </p>
$$U = \sum_{t=1}^{12} u(z_t).$$
We then assume that there exist a tax \(G\), which leads to \(U = \sum_{t=1}^{12} u(z_t - G_t)\).
-->
</div>
<hr>
<div id=future_research>
<h2>Future Research/Research Projects:
<a href="#future_research"></a>
</h2>
<hr style="height:1px;border:none;color:#333;background-color:#333;" />
<h4>Is France a land of opportunity? The Geography of Intergenerational Mobility in France.
</h4>
<ul compact>
<li>
<p>
<a href="https://academic.oup.com/qje/article/129/4/1553/1853754">Chetty & al(2014)</a> show that intergenerational income mobility is quite small in the US, moreover, they show that intergenerational mobility varies substantially across areas within the United States.
</p>
<p>
France intergenerational mobility has not been subject to many studies.
<a href="http://www.strategie.gouv.fr/sites/strategie.gouv.fr/files/atoms/files/note-36-geographie-ascension-sociale-ok.pdf">Dherbécourt (2015)</a> looked at the intergenerational mobility over life cycle with respect to socio-professional groups and shows great disparities between regions. However, since the information on parents and children income is missing, it does not allow to do a direct comparison with Chetty & al that rely on inter-quintile income mobility.
</p>
<p>
The Echantillon Demographique Permanent allows to observe over 70 000 children of 17 years old and the income they earn 6 years later. We can also observe their parents income while they live with them.
</p>
<p>
We can thus look at the relative position in the income distribution of the parents when their child is minor, then look at their children relative position in the income distribution 6 years later where most of them will have finished their studies and will have started to work.
</p>
<p>
The sample being quite large, with very precise geographical information it creates the possibility to investigate geographical intergenerational mobility at least on the county level.
</p>
<P>
These important descriptive statistics can then lead to a more thorough analysis by looking at the correlation in dimensions developed in Chetty & al. One of the dimensions that I am very interested in is the link between political color, contributions to public goods and social mobility.
</p>
<!--
is a set of administrative databases that allows to follow individuals in panel. If an other individual lives in the same household as an individual in the panel we can also observe their income. Consequently, we can observe parents income of a 17 years old child in the panel, and then observe the child income 6 years later. -->
</p>
</li>
</ul>
<hr style="height:1px;border:none;color:#333;background-color:#333;" />
<h4>Taxe Incentives and Fertility of Rich French Households:
</h4>
<ul compact>
<li>
<p></p>
France is facing a decline in its fertility rate since 2014. Fertility, and Tax credits and subsidies related to children has been quite stable from 2006 to 2014. Child per women has decreased from 2 in 2014 to 1.88 in 2018.<sup><a href="#fn1" id="ref1">1</a></sup> Over the same period, transfers towards households with children have changed: the total transfer per child has been increased for low-income households and decreased for high-income households.
</p>
<!--
<div class="embed-responsive embed-responsive-4by3">
<iframe class="embed-responsive-item" src="./bilan.html"></iframe>
</div> -->
<p>
Some commentators of the public debate claim a causal link between the decrease in child tax break and transfers towards high-income households and the global decline in fertility (<a href= "http://www.lefigaro.fr/vox/economie/2019/01/16/31007-20190116ARTFIG00295-comment-le-matraquage-fiscal-et-social-des-familles-a-fait-chuter-la-natalite-francaise.php?fbclid=IwAR00_d2LGT3yaifPtiKQQ71VwFiFgdczhGHa2TNr6jOlBCfEbI9szfPqQ6g">an example</a> -- in French).
</p>
<p>
While many studies have indeed found a positive link between tax exemption per dependent children and fertility (<a href="https://www.jstor.org/stable/pdf/2006683.pdf?seq=1#page_scan_tab_contents">Whittington & al (1990)</a>, <a href="https://www.tandfonline.com/doi/abs/10.1080/0032472031000150066?journalCode=rpst20">Gauthier & al(1997)</a>, <a href="https://www.mitpressjournals.org/doi/abs/10.1162/REST_a_00342">Cohen & al (2013)</a>, <a href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1157260">Laroque Salanie (2008)<a>, <a href="https://www.ntanet.org/wp-content/uploads/2018/03/Bastian_kidsandmarriageEITC_fulldraft-1.pdf"> Bastian(2017)</a>), these studies are usually made on the whole population or focus on the bottom of the income distribution.
<!--
Many other explanations can be made on the decline of France fertility (age of women at first child, economic uncertainty, provision of public goods such as creches or schools, etc). -->
</p>
<p>
Does this change in women fertility rate is due to the child tax break and subsidies directed towards high-income households?
To answer that question we can first rely on a relatively big database (the Échantillon Démographique Permanent) to describe from which part of the income distribution changes in fertility. We can include as a preliminary analysis the classical determinants of fertility such as the age of the mother at birth, the age of parents relation, number of years of studies, number of previous children, and so on, to see if indeed the decrease in French fertility is due to high-income households.</p>
<p> If the change in fertility is indeed due to change in high-income households, we can then answer the question if it is due to the reduction in child-related transfers and tax breaks. By relying on three French tax reforms that happened between 2012 and 2015 that reduced widely the tax exemption per dependent children (up to 6% of disposable income for a 3 children family)<sup><a href="#fn2" id="ref2">2</a></sup> only for the last decile of the income distribution. An evaluation of the impact of the reform could be done with a triple-difference estimator as used in the third chapter of my thesis. The Échantillon Démographique Permanent allows running such an analysis in a panel over a relatively large population (over 50 000 households).<br>
</p>
<hr></hr>
<sup id="fn1">1. <a href="https://www.insee.fr/fr/statistiques/3692693#graphique-figure3"> Bilan démographique 2018
La fécondité baisse depuis quatre ans (INSEE Première, 2019)</a>.<a href="#ref1" title="Jump back to footnote 1 in the text.">↩</a></sup> <br>
<sup id="fn2">2. <a href="./docs/IDEP/idep_analyses_N6.pdf"> Fiscalité des familles aisés : vers une forfaitarisation de l’enfant (RFFP, Bargain & al, 2016)</a>.<a href="#ref2" title="Jump back to footnote 1 in the text.">↩</a></sup>
<!-- https://sites.uclouvain.be/econ/DP/IRES/2016010.pdf -->
<!-- <a href= "http://budgetmodel.wharton.upenn.edu/issues/2016/1/7/the-economic-determinants-of-fertility-choices">click</a>-->
<!-- <p>
Top of the income distribution parents make their children inherit of their good social, and intellectual and capital. (Trzaskowski & al (2014))-->
<!-- https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3907681/ -->
<!--
They also invest more in their education. This would generally lead to more productive children.
By subsidizing rich households fertility government may create future economic growth when their high-skilled children will enter the labor market.
</p> -->
<!-- Although this reasoning can be questioned ethically, the government would in that case face a classical equity/efficiency tradeoff: by giving to rich households, the government entails the equity principle by giving more to those that are not in needs. But by incentivizing rich households to have children it may endogenously increase future growth through better human capital.
However, that debate is of interest <b>only on the condition</b> that "Rich Households" react on the fertility margin. Do they?
</p>
<!-- <p>
If they do not react, then child tax break increasing with income would have little moral grounds. And if they do react it opens room to
</p> -->
</li>
</ul>
<hr style="height:1px;border:none;color:#333;background-color:#333;" />
<h4>Temporality of the tax-benefit system, behavioral and normative implications:
</h4>
<ul compact>
<li>
<p>Based on the stream of income an individual has, she can, for the same amount of labor income, have different disposable income due to temporal consideration of the tax-benefit system. </p>
<!-- <p> A first example would be the French income benefit (RSA) that is based on the last three month preceding the claim for the benefit and is paid for three months after the claim. A direct consequence is that an individual that works every second month will have a lower disposable income than one that works every second quarter.</p> -->
<p> A first example would be the French income tax that is based on yearly income. Since it is progressive, a seasonal worker that work every year from October to March will pay a lower tax than an individual that works every second year.</p>
<p> A second example, a woman that give birth directly after graduating will have no income compensation for the potential loss (or opportunity cost) in labor income, while a woman that has been working for a long enough period will benefit from pre-maternal and maternal leaves.
</p>
<P>In these examples, the labor income earned could be the same over a given period (e.g. life cycle) but can lead to substantive differences in disposable income over that same given period based on the timing of income. <br>
</P>
<!-- <p>
This rise the question of whether or not it is desirable.
Vickrey has chosen to defend the point of view that it is not by stating that the same income over the life-cycle should lead to the same disposable income. </P> -->
<p> The French tax-benefit system has many complexities that lead to complex disincentives and incentives to work with respect to the timing of labor income interacting with the temporality rule of the tax-benefit system.
A precise simulation of the tax-benefits (ARE, RSA, PPA, but also in kind local benefits) would allow to simulate agents incentives to work. The use of administrative databases (CNAF data, FH-DADS) would allow seeing how those disincentives generated by the temporality of the tax-benefit system impact agents behavior. If it has an impact, it would be possible to compute the implied deadweight loss of such temporal disincentives to work.</p>
<p><b>Why is it important?:</b><br>
Technological advances make unqualified work less important in firms production functions. Labour laws generally encompass a minimum wage that may be above the marginal productivity of the lower-skilled workers, as consequence unemployment is important in many developed countries.
<br>
A direct consequence of unemployment is that low skilled workers usually have temporary contracts, that leads to highly volatile labor income.
</p>
<p>Tax-benefit systems of the developed country tend to have a regressive tax system when benefits are included, in the sense that the tax is concave at the bottom of the distribution. For instance, individual that earn less than the minimum wage in France face an effective marginal tax rate that is greater than 70%.</p>
<p>
These high marginal tax rates imply a very low incentive to work. A very small cost of job search or a small fixed cost of labor could easily make the expected gain of being active to be negative. Temporal considerations of the tax can increase or reduce these potential disincentives or incentives.
</p>
<p>The way we design taxes w.r.t tax temporality would impact thoroughly individuals at the lower end of the income distribution since their income varies a lot (see <a href="#tax_frequency">Tax frequency</a>), and that taxes represent a very significant share of their disposable income.
<p>
<p><b>Normative considerations:</b><br>
<p>
Luck egalitarianism being an appealing normative argument stating that individuals should be as well off independently of their unchose circumstances. Unemployment and temporary contracts are de-facto creating inequalities between individuals that are equally responsible for their situations. Thus a large part of the variation that the bottom of the income distribution faces can be considered as <b>brute luck</b>.
</p>
<p>
A solution to reduce the impact of taxation would be to use
<!-- <a href="./normative_tax_temporality_axioms.html"> </a>-->temporally neutral tax systems. For instance, Vickrey's tax system would be quite suitable.<br>
However, Vickrey's tax system would allow to do not respect the imperative that a tax should be paid by its citizen "in proportion to their respective abilities", allowing very productive individuals to work on a very small period of their lifetime by facing a very small tax.
</p>
<!-- Arneson and Cohen developed the concept of
opportunity-egalitarianism, where they argue that individuals should be
held responsible only for characteristics and decisions that are fully within their
control. -->
<!-- Dworkin (1981) has developed the normative arguments that individuals -->
</p>
<!--
<p>In the first chapter of my thesis, we determine two opposite normative views when designing taxes w.r.t temporality. -->
<!--
The goal of this project is to extend the analysis made in the first chapter of my thesis concerning the impact of the temporal choices made on tax deseign. </p> -->
</li>
</ul>
<!-- <h4> Version 2:
</h4>
<ul compact>
<li>
<p>
The rise of employment precariousness in developed countries, especially at the bottom of the income distribution, is logically associated with an increase in short-term labor income variations.
</p>
<p>
Infra-annual income variation of households is subject to very few studies in the economic literature. Hills (2006) shows that the very bottom of the distribution is the one that has the largest income variations within the year. They also show that the bottom of the distribution barely save, and usually just cope with their disposable income for a given month.
On another hand, taxes and benefits are linked with the timing of events, such as income realization, unemployment, pregnancy, etc.
</p>
<p>
Policy evaluations of the tax benefit system through microsimulation, but also optimal taxation models assume that individuals have steady incomes. A first issue, since tax-benefit schemes are not linear, can lead to radical differences between static simulations of tax and benefits throughout the year with a steady income, and the one households are really entitled to.
</p>
<p>
A second one is that optimal taxation models solutions are not valid anymore if we introduce within year income variations.
</p>
<p>
Moreover, this could make some benefits to create poverty traps that would not be observable with a static framework. Indeed the incentives usually based on a "work must pay" normative statement may not always hold when varying hours of work are introduced.
<p>
Vickrey (1939) proposed a potential solution for such issues that is based on income averaging over the life cycle. His proposition was that at each period of the life cycle, the cumulative tax paid is the one that one would have had to pay if its income were steady since the beginning of her life.
</p>
<p>
Vickrey's solution has very appealing features. Indeed if one has a low income at the beginning of her life, but then has later a positive productivity shock that leads to the possibility to earn a high income, the marginal income tax she will face would be low and will thus have desirable incentivizing properties. Conversely if one has a high income at the beginning of her life, but then has a negative productivity shock, she will receive negative tax transfers for multiple periods.
</p>
<p>
Obviously, there is no such thing as a free lunch, and Vickrey's solution comes with two major drawbacks. The first one is the other face of the coin of the preceding paragraph: if a high skill worker has a life accident that reduces her skill, she will face a very high marginal tax rate that in conjunction to the negative income tax she will get would lead her to do not work. (and it might lead to a potential deadweight loss of the tax greater than the one that one would have paid in the presence of an annual tax with a progressive scheme).
</p>
<p>
Another one concerns the normative statement that for a given income over the life cycle, one should pay the same tax whatever the timing of the income. The justification for a progressive tax scheme is normatively based on one ability to contribute. If one ability to contribute is very high at a given period, it can be viewed as fair that the burden of the tax is higher on the given period.
</p>
<p>
Vickrey's solution, however, does not take into account the responsibility of the tax-payer
</p>
<p>
The first task would be to document by making tests cases of different temporal situations with OpenFisca to document which cases violate the "work must pay" condition, but also document the actual incentives to work with respect to the timing of events. This would include national tax and benefits, but also local benefits (in money or in-kind). The analysis will be conducted for France, for localities for which local transfers has been coded (Paris, Le Mans, etc).
Once the best timing sequences have been correctly documented, it is possible, by using databases that contains infra-annual information on individuals income to what extent they are sensitive to the associated incentives. (Leila article chomage) Evaluate the induced deadweight loss of bad temporality (reference steady income).
Then, based on the risk that is inherent to being in a specific position of the income distribution, it is possible to derive
</p>
For instance, if an individual face a risk on its future employment status (that could be independent of her will)
Example of time violation:
https://content.iospress.com/articles/work/wor01645
https://www.sciencedirect.com/science/article/abs/pii/S0927537101000513
https://www.sciencedirect.com/science/article/abs/pii/S092753710500031X
</p> -->
<hr style="height:1px;border:none;color:#333;background-color:#333;" />
<h4>Other:
</h4>
<!-- <<<<<<< HEAD
In addition to those three projects, I also have a vivid interest for other research questions. I am very interested in gender inequalities. I am closely following the litterature on discrimination (either gender discrimination, race discrimination or social discrimination). I would also like to study couple formation and dissolution. I observed that the discontinuty at 50% observed in as Bertran et al is as important in the US as in France. I also saw on the cohabitants project that the fiscal gain to marriage are an important. I think that there is a room to study the interaction between gender norms,
======= -->
I also have a vivid interest in other research questions. I am very interested in gender inequalities. I am closely following the litterature on discrimination (either gender discrimination, race discrimination or social discrimination). I would also like to study couple formation and dissolution, notably with respect to the tax-benefit system incentives.
<!-- The discontinuty at 50% observed in as Bertran et al is as important in the US as in France. I also saw on the cohabitants project that the fiscal gain to marriage are an important. I think that there is a room to study the interaction between gender norms, -->
<!-- <p>To answer that question we are making the assumption that the yearly utility function is inter-temporally additive monthly function increasing in income \(z\): </p>
$$U = \sum_{t=1}^{12} u(z_t).$$
We then assume that there exist a tax \(G\), which leads to \(U = \sum_{t=1}^{12} u(z_t - G_t)\).
"Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum."
est laborum." -->
</div>
</div>
</div>
</div>
<iframe width="0" height="0" src="https://www.youtube.com/embed/WjG2R8S9lGk?autoplay=0;start=3"
frameborder="0" allow="autoplay; encrypted-media" allowfullscreen>
</iframe>
<div class="jumbotron text-center" style="margin-bottom:0">
<p>Mail:
<a href="mailto:[email protected]"><span class="far fa-envelope"></span></a>
</p>
<!-- <address>
Written by <a href="mailto:[email protected]">Jon Doe</a>.<br>
Visit us at:<br>
Example.com<br>
Box 564, Disneyland<br>
USA
</address> -->
<p>
Twitter: <a href="https://twitter.com/psyfico">
<span class="fab fa-twitter"></span>
</a>
GitHub: <a href="https://github.com/adrienpacifico">
<span class="fab fa-github"></span>
</a>
StackOverflow: <a href="https://stackoverflow.com/users/5304366/adrien-pacifico">
<span class="fab fa-stack-overflow"</span>
</a>
</p>
</div>
</body>
</html>
<!--
<h4>Taxe Incentives and Fertility of Rich French Households:
</h4>
<ul compact>
<li>
<p>
Many studies have found a positive link between tax exemption per dependent children and fertility (<a href="https://www.jstor.org/stable/pdf/2006683.pdf?seq=1#page_scan_tab_contents">Whittington & al (1990)</a>, <a href="https://www.tandfonline.com/doi/abs/10.1080/0032472031000150066?journalCode=rpst20">Gauthier & al(1997)</a>, <a href="https://www.mitpressjournals.org/doi/abs/10.1162/REST_a_00342">Cohen & al (2013)</a>, <a href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1157260">Laroque Salanie (2008)<a>, <a href="https://www.ntanet.org/wp-content/uploads/2018/03/Bastian_kidsandmarriageEITC_fulldraft-1.pdf"> Bastian(2017)</a>) suggesting that children are normal goods. However, these studies are usually made on the whole population or focus on the bottom of the income distribution.
</p>
<!-- https://sites.uclouvain.be/econ/DP/IRES/2016010.pdf -->
<!-- <a href= "http://budgetmodel.wharton.upenn.edu/issues/2016/1/7/the-economic-determinants-of-fertility-choices">click</a>-->
<!--
Proponents of child tax break increasing with income may argue on the possibility that, by incentivizing rich households to have more children may endogenously increase future growth through better human capital. -->
<!-- <p>
Indeed top of the income distribution parents tend to make their children inherit of their good social, and intellectual capital.
They also invest more in their education. This would generally lead to more productive children.
By subsidizing rich households fertility government may create future economic growth when their high-skilled children will enter the labor market.
</p> -->
<!--(Trzaskowski & al (2014))-->
<!-- https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3907681/ -->
<!-- <p>
Although this reasoning can be questioned ethically,
the government would then face a classical equity/efficiency tradeoff: by giving to rich households, the government entails the equity principle by giving more to those that are not in needs. But by incentivizing rich households to have children it may endogenously increase future growth through better human capital.
As child tax break positively linked with income being generalized in developed countries, the consequences and normatives choices implied needs to be analysed.
this debate is of interest to the condition that rich households do react to fertility incentives.
</p> --> -->