NCEI versus PSR compared

For many years I have studied on and off the NOAA US divisional statistics of temperature and precipitation, because climate statistics — like all statistics — are always interesting to me if at times capable of making me very angry when runaway climate change is shown.

What is interesting is that NOAA has two sets of precipitation (and temperature) sets for each of the contiguous US states. One is published by the National Center for Environmental Information and the other by the Physical Sciences Laboratory. Over the period I have used them to study climate, I have assumed the two to be essentially completely interchangeable for the purpose of statistical tables, although for graphical purposes I can prefer one or the other (I prefer NCEI for precipitation maps and PSL for temperature maps) depending on the precise maps offered.

When I realised NCEI was offering comparisons for all months, I decided to try to look at each continental state (this excludes Hawaii) and tabulate its wettest and driest month since record began. However, when I tried to do this on the PSL site, I found that there appeared to be some differences between the two. To test this, I tabulated the wettest and driest months for all states east of the Continental Divide. For states west of the Divide there are generally sufficient months with zero or negligible precipitation that I decided only to tabulate each state’s wettest month. I have included both sets of units for convenience of readers not familiar with US units.

State	Wettest month						Driest month (omitted for states west of Divide)
	NCEI			PSL			NCEI			PSL
	Month	in	mm	Month	in	mm	Month	in	mm	Month	in	mm
Alabama	July 1916	17.59	446.8	July 1916	18.13	460.5	October 1963	0.03	0.8	October 1963	0.03	0.8
Alaska	September 1925	7.59	192.8				February 1979	0.72	18.3
Arizona	January 1993	5.22	132.6	January 1993	5.72	145.3
Arkansas	October 2009	14.13	358.9	October 2009	13.38	339.9	October 1963	0.13	3.3	October 1963	0.18	4.6
California	January 1995	12.50	317.5	January 1909	13.50	342.9
Colorado	April 1900	5.45	138.4	April 1900	5.13	130.3	November 1904	0.03	0.8	November 1904	0.03	0.8
Connecticut	October 2005	15.92	404.4	October 2005	16.02	406.9	March 1915	0.28	7.1	March 1915	0.27	6.9
Delaware	August 1967	12.58	319.5	September 1935	11.50	292.1	October 1963	0.12	3.0	October 1963	0.13	3.3
Florida	August 1898	13.49	342.6	September 2004	14.25	362.0	November 2016	0.21	5.3	October 1940	0.16	4.1
										November 2016
Georgia	July 1916	13.46	341.9	July 1916	14.06	357.1	October 1963	0.06	1.5	October 1963	0.04	1.0
Idaho	December 1964	8.52	216.4	December 1964	6.72	170.7
Illinois	September 1926	9.62	244.3	September 1926	10.18	258.6	October 1964	0.18	4.6	October 1964	0.15	3.8
Indiana	January 1950	10.11	256.8	January 1950	10.01	254.3	March 1910	0.21	5.3	March 1910	0.22	5.6
Iowa	July 1993	10.45	265.4	July 1993	10.75	273.1	October 1952	0.02	0.5	October 1952	0.02	0.5
Kansas	May 2019	10.51	267.0	May 2019	10.22	259.6	January 1986	0.00	0.0	January 1986	0.00	0.0
							November 1989	0.00	0.0	November 1989	0.00	0.0
Kentucky	January 1937	15.16	385.1	January 1937	15.49	393.4	October 1963	0.14	3.6	October 1963	0.14	3.6
Louisiana	May 1907	15.09	383.3	October 2009	16.41	416.8	October 1952	0.03	0.8	October 1952	0.03	0.8
Maine	October 2005	11.26	286.0	October 2005	12.62	320.5	March 1915	0.46	11.7	October 1947	0.37	9.4
Maryland	August 1955	11.85	301.0	July 1945	12.68	322.1	October 1963	0.15	3.8	October 2000	0.12	3.0
Massachusetts	October 2005	14.03	356.4	October 2005	13.99	355.3	March 1915	0.09	2.3	March 1915	0.11	2.8
Michigan	September 1986	8.72	221.5	June 1943	7.22	183.4	March 1910	0.26	6.6	March 1910	0.22	5.6
Minnesota	June 2014	8.07	205.0	June 2014	8.13	206.5	December 1913	0.07	1.8	December 1913	0.06	1.5
Mississippi	April 1991	14.36	364.7	December 1982	15.62	396.7	October 1963	0.07	1.8	October 1963	0.05	1.3
Missouri	September 1993	11.31	287.3	June 1928	11.69	296.9	January 1986	0.10	2.5	January 1986	0.21	5.3
Montana	June 1944	6.38	162.1	June 1944	6.40	162.6	October 1987	0.11	2.8	October 1987	0.12	3.0
Nebraska	June 1967	8.20	208.3	June 1967	8.64	219.5	November 1914	0.01	0.3	November 1914	0.01	0.3
										November 1939	0.01	0.3
Nevada	January 1916	3.65	92.7	January 1916	3.63	92.2
New Hampshire	October 2005	13.65	346.7	October 2005	13.23	336.0	March 1915	0.19	4.8	March 1915	0.22	5.6
New Jersey	August 2011	15.67	398.0	August 2011	15.79	401.1	October 1963	0.25	6.4	October 1963	0.25	6.4
New Mexico	September 1941	5.48	139.2	July 1911	4.60	116.8
New York	June 1972	8.58	217.9	October 2005	9.38	238.3	October 1963	0.34	8.6	October 1963	0.41	10.4
North Carolina	September 1999	13.22	335.8	August 1901	13.00	330.2	October 2000	0.08	2.0	October 2000	0.06	1.5
North Dakota	July 1993	7.97	202.4	July 1993	8.11	206.0	November 1939	0.03	0.8	November 1939	0.03	0.8
Ohio	January 1937	9.69	246.1	January 1937	9.42	239.3	October 1963	0.23	5.8	October 1963	0.24	6.1
Oklahoma	May 2015	14.44	366.8	May 2015	14.40	365.8	January 1986	0.03	0.8	January 1986	0.04	1.0
Oregon	December 1964	13.05	331.5	December 1964	16.33	414.8
Pennsylvania	June 1972	11.22	285.0	June 1972	11.63	295.4	October 1963	0.16	4.1	October 1963	0.22	5.6
Rhode Island	March 2010	14.75	374.7	March 2010	14.76	374.9	March 1915	0.14	3.6	March 1915	0.14	3.6
South Carolina	July 1916	14.41	366.0	August 1901	23.60	599.4	October 2000	0.02	0.5	October 2000	0.07	1.8
South Dakota	May 1962	6.78	172.2	May 1962	6.88	174.8	November 1939	0.01	0.3	November 1939	0.01	0.3
Tennessee	January 1937	14.53	369.1	January 1937	13.41	340.6	October 1963	0.03	0.8	October 1963	0.03	0.8
Texas	May 2015	9.05	229.9	May 2015	11.15	283.2	October 1952	0.03	0.8	October 1952	0.02	0.5
Utah	October 1972	3.69	93.7	October 2004	4.10	104.1
Vermont	November 1927	10.61	269.5	November 1927	10.48	266.2	March 1915	0.32	8.1	March 1915	0.35	8.9
Virginia	September 1999	13.84	351.5	September 1999	10.37	263.4	October 2000	0.09	2.3	October 2000	0.03	0.8
Washington	December 1933	16.61	421.9	December 1933	24.36	618.7
West Virginia	November 1985	9.53	242.1	September 2018	9.48	240.8	October 1963	0.24	6.1	October 1963	0.23	5.8
Wisconsin	September 1986	8.38	212.9	September 1986	8.43	214.1	February 1958	0.11	2.8	February 1969	0.10	2.5
							February 1969	0.11	2.8
Wyoming	May 1908	5.09	129.3	May 1908	5.83	148.1	November 1939	0.02	0.5	November 1939	0.02	0.5

If we look closely, we see that there are a few cases where NCEI and PSR do differ substantially (shaded in red). Almost all of these are for states’ wettest months — the driest months vary less between the two sets and there are many fewer months listed even comparing only the states east of the Divide.

In the cases of Washington state and Idaho, whilst the identity of the state’s wettest-ever month is definite, the estimated actual amount of precipitation varies by as much as a quarter.

In the cases of California and West Virginia, the precipitation figures for the state’s two wettest months are very nearly equal and which is the wettest may not be perfectly clear.

The reason for these discrepancies if not perfectly clear. In the late 2000s, state precipitation measurements were significantly updated to take into account much higher precipitation at high altitudes (similar updates occurred for Tasmania in Australia, and for the United Kingdom). Differences in this update for mountainous regions may be important, but a variety of other factors like:

1. difficulties measuring tropical cyclone rainfall in the South Atlantic States and New York
2. difficulties measuring lake-effect precipitation in Michigan

may also be important explaining the observed differences. I have not yet researched possible causes of the differences between NCEI and PSL, but still find it interesting to note that they exist.

Racism, labour supply shocks and the contrasting responses of the US and South Africa

In last year’s Shocking Contrasts: Political Responses to Exogenous Supply Shocks, Ron Rogowski discussed a number of supply shocks to land, labour and human capital. Here I noted that he never discussed a case involving exogenous gain of land, nor of loss or gain of physical capital. My older post was itself about the effects of sudden gain of land, and I do wish to discuss the physical capital issue, but will do so later. More interesting is that, whilst Rogowski discusses loss of labour through high mortality and loss of land via declining trade he never discusses loss of labour in a land-rich society through the latter mechanism. I have noted that this could have profound effects before here and here. For a labour-scarce twentieth-century nation, complete loss of access to the abundant labour of Europe and monsoon Asia would have created a much larger supply shock than the Black Death. Even in the populous United States, it would have amounted to a labour supply cut of 95 percent by a crude calculation, and elsewhere it would have been larger still. Complete loss of access to the labour of Europe and monsoon Asia would have totally upended the racial hierarchies of the US and South Africa. Oppressed black populations would have used skyrocketing wages from a decimated effective labour supply (noted in the first paragraph) to demand political and economic rights, whilst the wealthy would have been literally crippled because of:

• limited latitude to substitute for labour due to already having done so substantially due to labour’s scarcity
• risk of further retaliation from abroad if the ruling classes tried to further restrict black mobility or wages
• risk of disloyalty from their poor white supporters if they attempted a military solution
• this was a major factor behind Goldwater’s 1964 landslide loss, as many racially conservative whites refused to risk his Vietnam policies involving possible nuclear strikes

Shocking Contrasts never uses Southern Africa’s loss of access to the highly abundant labour of South Asia as an example of a supply shock. In communication with me Rogowski argued that it was endogenous and anticipated, but I question his argument. Before Indian independence similar if less severe racial policies had no effect. South Africa thus gained unlimited access via trade to labour-intensive products from South Asia. Hence, even if incoming Prime Minister François Daniel Malan and his advisers knew changes were likely when South Asia became independent, they would have not assumed major consequences.

The United States had similar laws of racial discrimination that effectively excluded African–Americans from voting or from living in virtually any rural or suburban area outside the plantation South, despite constitutional amendments aimed at creating such rights. Expanding trade after 1870 weakened the economic position of black Americans and increased the determination of whites to eliminate theoretical constitutional rights. The position of black Americans steadily deteriorated and — as had been noted by many commentators from Walter Dean Burnham to Frances Fox Piven and would occur again after 1973 — political influence was monopolised by the rich while poorer whites voted in smaller and smaller numbers. (Not discussing why this happened at the very time political participation in Europe was rising and suffrage expanding was a major omission from Commerce and Coalitions, although the contrast is precisely analogous to Eastern Europe versus Western Europe in the long sixteenth century discussed on pages 157 and 158.)

Mass influence upon policy in trade-open labour-scarce (as the US and South Africa are) economies is exceedingly unlikely for the simple reason that free trade eliminates requisite economic power for all except minimally the top decile and perhaps an even smaller elite. Nonetheless, neither:

1. those who acknowledge that mass interest groups have zero influence upon US government policy and that this state of affairs has been maintained since the 1973 oil crisis, nor
2. writers like Rogowski of David Lektzian and Dennis Patterson (who in the 2000s analysed theoretical practical effects of sanctions on countries with varying factor abundances)

seriously consider extremely plausible arguments that:

1. mass influence on politics in trade-open labour-scarce economies is generally implausible
2. the most probable political models for a trade-open labour-scarce society are the family business monarchy of the Gulf States or the semi-family dictatorships of the East Slavic and Turkic “core USSR”
3. only collapse of global trade could potentially produce anything resembling democratisation of such societies

By the 1940s, the US ruling class was acquainted with the spectre of widespread worker revolution in Europe. If successful, US capitalists would have been placed in direct conflict with workers’ states dedicated at least theoretically to their expropriation, and likely to redirecting exports away from the US. Another threat, already developed to some extent during the interwar period, was restriction of foreign access to the products of abundant local labour by bourgeois nationalists who wanted independent local development.

The US ruling class understood it had to compromise with Europe and newly independent Asian nations to expand trade. However, given that it gained immense profits by oppressing blacks, it never assumed that greater rights and opportunities for the mass of African-Americans would remain permanent. US policies between Smith v. Allwright and the 1973 oil crisis were contradictory because of conflicting needs to appease European workers and nonwhite nations versus maximising the profits and power of the rich. Rogowski notes that black leaders did not pursue the protectionist (or anti-IMF and anti-GATT) policies that would have immensely benefitted the mass of African-Americans, and their reliance on a federal government with conflicting needs is almost certainly the reason. Once the 1973 oil crisis established an alternative rentier-state model that made tax revolts by the super-rich easier and more rewarding, the US ruling class could establish a much more consistent policy of reversing civil rights gains, lowering or eliminating taxes on the rich, and completely eliminating the influence of mass interest groups (the bottom ninety percent) on government policy.

Unlike the US ruling class and perhaps even its white supporters, the National Party of South Africa and its Boer supporters undoubtedly came out of the war with no understanding of either the newly independent nations of Asia or the newly aligned nations of the Soviet bloc. Consequently, South Africa would be exposed to the labour supply shock that the US ruling class avoided. Its black population could therefore fight a longer and more effective struggle against apartheid. Unlike Jim Crow, apartheid would be completely destroyed, and with it South Africa‘s white ruling elite.

One might think the difference resulted from the much greater resources of the US ruling class, which gave it more access to information about global geopolitics. However, even under apartheid South Africa was not isolated from geopolitics and was relied on to restrain Stalinism in Africa. Moreover, Jan Smuts, Malan’s long-time predecessor, had played a key role in establishing the League of Nations, placing even more doubt that South African elites were that ignorant. Another critical difference might be that the most insular (and racially intolerant) regions of the United States — the Deep South, the Great Basin and the Plains — were virtually unrepresented amongst the postwar political and judicial elite. Thus, the most likely reason for the difference is that the US ruling class was much more acquainted with the threat of Communism and thus invested in short-term domestic compromises to help defeat it.